Disclaimer

No part of this material may be reproduced, stored in a retrieval system, or

transmitted in any form by any means, electronic or mechanical including

photocopying, recording or by any information storage and retrieval system,

without prior written permission of the publisher.

The opinions expressed and figures provided in this proceedings of CAIRIA-2018

are sole responsibility of the authors. The organizers and the editor bear no

responsibility in this regard. Any and all such liabilities are disclaimed.

Print: ISBN 978-93-5300-754-6

Copyright © CAIRIA-2018 All Rights Reserved. Printed in INDIA. March 2018

Published and Printed by Publication Committee Of Conference.

Designed by Ayush Kumar Rathore BSc.(IT)

C A I R I A - 2 0 1 8

C O N F E R E N C E O N A R T I F I C I A L I N T E L L I G E N C E :

R E S E A R C H , I N N O V A T I O N & I T S A P P L I C A T IO N S

1 4 - 1 5 M A R C H - 2 0 1 8

AMITY INSTITUTE OF INFORMATION TECHNOLOGY,

AMITY UNIVERSITY, LUCKNOW CAMPUS

Proceedings

Conference on Artificial Intelligence: Research, Innovation & its Applications (CAIRIA-2018)

Event Photographs

Keynote Speakers: Dr. Laxmidhar Behera, Professor, Indian Institute of Technology, Kanpur; Dr. Rohit Mehrotra,

Surgeon, GSV Medical University, Kanpur; Along with Professor STH Abidi; Prof. S. K. Singh; Mr. Deepak

Sharma, Chairman CSI; Brig. U. K. Chopra; Wg. Cdr. (Dr.) Anil K. Tiwari & Dr. Pankaj K. Goswami.

Keynote Speakers: Dr. Laxmidhar Behera, Professor, Indian Institute of Technology, Kanpur; Dr. Vivek Singh,

Professor, IIT BHU; Dr. Rohit Mehrotra, Surgeon, GSV Medical University, Kanpur; Along with Conference

Organizing Committee.

Keynote Speakers: Dr. Nischal K. Verma, Professor IIT-Kanpur & Mr Shivank Shekhar, Co-Chair, Global

WebVR Industry Committee, VRAR Association Along with Professor S. K. Singh; Dr. Parul Verma;

Ms Meenakshi Srivastava; Mr. Sanjay Taneja; Dr. Pankaj K. Goswami and participants.

Brig. Umesh K. Chopra (Retd.) Director-AIIT,

Amity University Lucknow Campus

Message

It is a matter of esteemed privilege to be part of the Conference on Artificial Intelligence: Research, Innovation & its Applications (CAIRIA-2018) being organized by Amity Institute of Information Technology, Amity University Lucknow Campus. The theme of the conference is of great importance and extreme relevance. It will focus on technology innovation and trend setting initiatives for Academia, Research & Development and Healthcare domain. Technology is moving ahead from digital age to the smart age of computing. An Artificial Intelligence is one of the most scope-worthy areas and also enhances the great value along with Robotics, Machine Learning, Machine Translation, Neural Networks and Fuzzy Systems. Artificial Intelligence techniques in healthcare have led to early detection of diseases including Cancer and even help in treatments. This Conference will focus on technological innovation and trend setting initiatives for the purpose and objectives of transforming the society into a knowledge base community. Artificial Intelligence researchers have created many tools to solve the most difficult problems being faced by society. I extend my sincere thanks to the organizing-committee for the success of the conference.

Brig. Umesh K. Chopra (Retd.)

Sanjay Mohapatra President

Computer Society Of India,

Mumbai

Message

It gives me great pleasure to know that Amity University Uttar Pradesh Lucknow Campus in association with Lucknow Chapter of Computer Society of India & IET is organizing a National Conference CAIRIA-2018 on " Artificial Intelligence : Research, Innovation & its Applications " on 14th & 15th March, 2018 at Lucknow.

This is the need of the present India and a right direction to look, when the country is

rising to be a global player and trying to improve the quality of life of its citizen. I am sure that this Seminar shall be a step forward and contribute in providing an insight towards the developments in Artificial Intelligence and will contribute in fulfilling dream and mission of 'Digital India'.

I send them my Best Wishes for the success of the Conference and look forward to the

ideas and the knowledge; the Conference is going to generate, which will help India and the Lucknow region, in realizing its true potential.

(Sanjay Mohapatra)

President, CSI

Advisory Committee

Dr. Y.N. Singh Professor, IIT Kanpur

Dr. S.K. Dwivedi

Professor, BBA, University, Lucknow

Dr. O.P. Sharma Associate Professor, IIT Patna

Mr. Deepak Sharma

Chairman, CSI Lucknow Chapter

Mr. Arvind Sharma Regional Vice President, Region-1, CSI Mumbai

Dr. Geetika Srivastava

Assistant Professor, Amity University

Organizing Committee

Chief Patron Dr. Aseem Chauhan Chairman, AUUP Lucknow Campus, Chancellor, Amity University, Rajasthan Patron Maj. Gen. K.K.Ohri AVSM (Retd.) Pro-Vice Chancellor, AUUP Lucknow Campus Director of the conference Brig. Umesh K. Chopra (Retd.) Director, AIIT, AUUP Lucknow Campus Chair Professor Dr. S. K. Singh Prog. Director, AIIT Mob. : 9415617702

Conference Chair Dr. Pankaj K. Goswami Mob. : 9453434364 Publication Chair Dr. Parul Verma Mob. : 9839289870 Finance Chair Mr. Sanjay D. Taneja Mob. : 9415874468, Registration Conduct Chair Ms. Namrata Nagpal Mob. : 9838930104, Technical Program Committee (TPC) Chair Ms. Meenakshi Srivastava Mob. : 9415433194,

Organising Committee ( Students )

Mr. Nikhil Singh Rawat

8317032920

Mr. Rishab Nigam

7068832340

Mr. Ayush Kumar Rathore

7905180613

Mr. Vinayak Singh

7860644732

Mr. Devanshu Srivastava

9129375459

Mr. Jasmeet Singh

7376263516

Mr. Abhishek Soni

9559778090

Mr. Varun Singh

7668721334

Mr. Prafulla Sahu

7275232428

Mr. Dravin Palekar

7007347310

Ms. Jigeesha Agarwal

7309030817

Ms. Sonal Jaiswal

7408594979

Abstracts

Nishchal K. Verma, Ph.D. (IIT Delhi) Devendra Shukla Young Faculty Research

Fellow, 2013-16, Associate Professor,

Department of Electrical Engineering, Indian Institute of Technology, Kanpur

Abstract

Nowadays, the data generation is increasing at exponential rate day by day. With huge size (terabytes and petabytes) of data available, big data brings large opportunities and transformative potential for public and private sectors. But, handling such huge amount of data using conventional machine learning is difficult. Deep learning techniques are the advancement in machine leaning is quite popular due to their adaptive nature for handling such type of data. But in real world applications, data generated exists in both linguistic and numeric form. Extracting relevant information from such types of data using deep learning is difficult. The deep fuzzy network (DFN) which is advancement in deep neural network (DNN) combines both linguistic and numeric form of data to transform the data into relevant information. This lectures starts with basic introduction to deep neural networks (DNN) and deep fuzzy networks (DFN). Further these networks are used to extract useful representations from the data for various applications i.e. machine health monitoring data, gene expression microarray data etc.

Dr Vivek Singh Professor, Computer Science,

Banaras Hindu University, Varanasi

Abstract

The work in Artificial Intelligence (AI) since 1956, when the term was first coined by John Mc Carthy in the Dartmouth summer research project on AI, has witnessed cycles of successes, misplaced optimisms & resulting cutbacks in enthusiasm and funding. While every small success aroused further interest and enthusiasm; major theoretical & experimental failures resulted in the research funding being reduced drastically. Most of the initial work on AI was concerned with one of the three areas: Language translation, Problem solving and Pattern recognition. There were noticeable successes, although at a micro scale, with Newel & Simon’s Logic Theorist (followed by GPS) and systems to handle Morse code & deciphering simple patterns along with important works of Marvin Minsky, Warren Mc Culloh & Walter Pitt etc. It was widely thought that as soon as cheap & faster hardware and memories are available, the simple theories and successes could be scaled up to work on bigger problems. However, most of the early systems which performed well on simple problems, turned to fail miserably when tried out on bigger and more difficult ones. This forced AI researchers to have a relook at the approaches and the philosophical directions.

Dr. Rohit Mehrotra Professor,

GSVM University, Kanpur

Abstract

Over 5% of the world’s population – or 466 million people – have disabling hearing loss (432 million adults and 34 million children). It is estimated that by 2050 over 900 million people – or one in every ten people – will have disabling hearing loss. 5-6 in every 1,000 children born in India are deaf. 45000 children born every day in India so about 250 deaf children born each day. If a Child can’t hear, how can the Child learn words & language? So, he can’t speak. Etiology of hearing loss in children remains obscure in about half of the cases. We can now check the hearing of a new born baby on the day of birth by OAE and BERA Test. On testing if child is deaf and is operated in the 1st year of life, the child hears and speaks normally. Best results are with bilateral implantation. If these surgeries are done in time there would not be any deaf and dumb in our society. But if surgery is delayed results start reducing and surgery does not benefit after 5 years of age. Early intervention is the key to success. This life transforming surgery of Cochlear Implantation gives normal hearing and speech. So with the aim of deaf free UP by 2030, the mission is to give hearing to every deaf child. Artificial Intelligence can be used to make speech processing faster and better, implant smaller and completely implantable.

Mr. Shivank Shekhar Co-chair

Global WebVR, Industry Committee,

VRAR Association

Abstract

Coalescing spectrums of AI & XR Once Oracle CEO Mark Hurd wrote on LinkedIn that almost three-quarters of companies respond to digital disruption only after the second year of its occurrence, while only 14% of executives believe their companies are ready to effectively redesign their organizations. So in the corporate world, there is a huge window of opportunity when it comes to understanding and applying the underlying technologies here. Congruence of the applicability of VR and Artificial intelligence excites me to introduce students to the frontline research and industrial application which have already been done so far.

Papers

Principle and Applications of Speaker Recognition

Security System Nilu Singh

1, Alka Agrawal

2 and R. A. Khan

3

Chennupati Jagadish1, Chris Bailey

2, and Parviz Famouri

3

1,2,3

SIST-DIT, Babasaheb Bhimrao Ambedkar University (A Central University), Lucknow, UP, India

Email: nilu.chouhan@hotmail.com

Abstract — This paper overviews the principle and

applications of speaker recognition. Speech is a natural way to

convey information by humans. Speech signal is enriched with

information of the individual. To recognize a person by his/her

voice is known as speaker recognition (SR). Since human voice

has some measurable characteristics hence it falls in the

category of biometric. The term biometric is used to measure

human’s body related characteristics. Biometric is also known

as realistic authentication. Voice biometric or speaker

recognition is used to recognize an individual through their

voice’s individual characteristic. Index Terms – Speaker Recognition, Speaker Verification,

Speaker Identification Open-set & Closed-set, Applications of

Speaker Recognition.

I. INTRODUCTION

Human speech is a medium for expressing their thoughts

during communication. Voice is a medium for human to

express their thoughts and information. A speech signal is a

complex signal which is packed with several knowledge

resources such as acoustic, articulatory, semantics, linguistic

and many more [1-2]. During communication, human easily

understand information such as emotion, language, and

mental status etc. This ability of human to decode

information motivated researchers to cognize speech signal

related information. And this idea helps to emerging a

system which able to procure and process the assembled

information of a speech signal. A person‟s voice is different

from another due to the acoustic properties of speech signal.

Speaker‟s voice is unique to an individual due to differences

which occur as structure of glottal and dissimilarities in the

vocal tract and the cultured speaking behaviors of

individuals [1-3].

In this digital era speaker recognition is the most useful

biometric recognition technique [3]. Now days many

organizations like bank, industries, access control systems

etc. are using this technology for providing greater security

to their vast databases [3-4]. Speaker recognition is mainly

divided into speaker identification (1: N matching) and

speaker verification (1:1 matching). Identification is

considered as more difficult than verification [5]. This is

intuitive that performance of speaker identification system

affected by the number of enrolled speakers (the probability

of incorrect decision increases). While the performance of

speaker verification system is not affected by increase in

voice database size since only two speakers are compared.

In last few years, requirement for authentication has

been increased with the increasing digital world of

information. It has already been proved that a biometrics

authentication technique increases security levels. Speaker

identification is a procedure of recognizing an utterance from

the enrolled speakers while speaker verification is a binary

task that is either speaker accepted or rejected. Speaker

verification systems are the real example of biometric

authentication systems. Further, it can be categorized as text-

dependent and text-independent [6]. The text-dependent

systems are based on same utterance spoken by speaker in

both cases i.e. training and testing while in text-independent

systems it is not required to utter the same sentence/words

during training and testing [5]. It is accepted that text-

dependent systems provide more precise results as both the

content and voice can be matched that is speaker utters exact

the sentence which he/she uttered during training. While

text-independent recognition systems, either use the same

voice sample or different for verification/identification [7-8].

Gaussian mixture model (GMM) is one of the most

common method used to speaker modeling for speaker

identification. GMM is used as two distinct ways for

identification system; firstly, when training database

principle is the maximum likelihood (ML) and parameter

estimation is performed by using expectation maximization

(EM) algorithm; and secondly when the training database

principle is maximum a posteriori (MAP)[9-11].

Speaker verification is used for those applications

where speech is used as main component to authorize the

speaker. The task of speaker identification is to decide that a

given utterance comes from a certain registered speaker.

Speaker verification usability is more than speaker

identification. With the increase in voice database, difficulty

of speaker identification increases. Speaker verification is

independent of voice database population since it works only

on binary decision that is acceptance or rejection. Speaker

recognition system performance (recognition accuracy) is

most affected by intersession variability (variability over

time) and spectra of a speakers speech signal [7] [12].

II. SPEAKER RECOGNITION

It is well accepted that in this electronic era people interact

using voice with the help of electronic devices. Human voice

is a signal which contains several information related to

human characteristics, such as emotion, words, language,

speaker identity etc.. To identify a human being by their

voice, required speech features are selected from speech

signal with available feature extraction techniques [13]. It is

the process of recognizing a person on the basis of his/her

voice. Automatic speaker recognition system is categorized

as speaker identification and speaker verification. Speaker

verification system decision is binary i.e. 0 or 1 (accept or

reject) as this justifies an identity claimed by the speaker.

Speaker identification decision requires N matching and then

the decision is made about acceptance or rejection [14].

Further it is distinguished as text-dependent and text-

independent speaker recognition. In the text-dependent

system, the recognition of speaker‟s identity is depends on

the specific words or sentences. In the case of text-

independent recognition, speakers have no restriction to

speak sentence or phrases [3] [15-19]. Fig. 1, presents an

overview of speaker recognition system and Fig. 2, shows

the basic concept of speaker identification and speaker

verification methodology.

Fig. 1 Basic speaker recognition system

The popularity of speaker recognition system is due

to its low cost of implementation. This is because of the

easily availability of microphones and the universal

telephone network. As in this digital era it is very easy to

capture someone‟s voice and authenticate it by using

speaker recognition system. The only cost is due to the

software which is used for speaker recognition system. The

study of speech signal is about its characteristics which

distinguish one speech signal with another [3] [15] [17].

Fig. 2 Speaker identification and speaker verification

A. Speaker Identification

Speaker identification system is 1: n matching

system. In this, user needs not to provide his/her identity to

the system. During identification user has to input his/her

speech to the ASR system and the system now decides the

identity of user on the basis of the match score. In this case

system has to perform N comparison (N is the number of

stored speaker/user model of voice database). During

identification, comparison with each registered model will

produce a likelihood score, on the basis of this score higher

likely model is identified for the speaker [7] [20].

From the study it is clear that speaker identification

is complex than speaker verification. Hence in case of

speaker identification, system performance degrades as

compared to speaker verification [21]. Fig. 3, shows the

process of speaker identification system.

Fig. 3 Process of speaker identification system

B. Speaker Verification

Speaker verification system is 1:1 that is the system

either accepts or rejects. In verification process, firstly user

need to provide his/her identity and then it is checked by the

system and decisions are made accordingly that whether the

claimed identity is true (accept) or false (reject) [22].

Speaker verification can be explained easily with the help of

an example of Automated Teller Machine (ATM). Before

any transaction, users are first needed to insert the ATM card

in the machine. This credit/debit card contains the

information about user such as name, signature etc. Now, if

the ATM is working on ASR technology then it will check

that the card is used by its genuine holder by asking to

produce his/her voice. Since the user has already provided

his/her identity to the system, only „yes or no‟ decision has to

be made by the ATM machine i.e. either the card holder is

accepted or rejected. This decision is made on the basis of

comparing the voice input to the previous voice input

provided by the user [2] [5]. Fig. 4 shows the process of

speaker verification system.

Model for each

Speaker

θ

C. Open-Set vs. Closed-Set

Speaker identification is categorized as closed-set and

open-set. Open-set speaker identification for a test utterance

of enrolled speakers is a twofold problem. Primarily, it is

necessary to find speaker model which have best matches in

the given set; Furthermore, it must be determined that the

best match test utterance is actually formed by the best

matched model speaker or some unknown speaker [2] [21]

[23-24]. Fig. 5 shows the classification of speaker

recognition system.

The possible error and problems in open set speaker

identification can be examined as follows. Suppose that N

speakers are enrolled in the system for voice database and

M1, M2……Mi, are their statistical models [25]. If O

represents the feature vector sequence extracted from the

given utterance, then the open-set identification is given as

follows:

Where θ is a pre-defined threshold and O is assigned to

speaker model. If the maximum likelihood score is greater

than the threshold θ, it means the voice is originated from a

known speaker. For a given O following errors are possible

[2] [23][26].

False Acceptance (FA): The recognition system

accepts a pretender as one of the registered speaker

[2].

False Rejection (FR): The recognition system

discards a registered speaker [2].

Speaker Confusion (SC): The system properly

accepts a registered speaker but demented with

another registered speaker [2].

Fig. 5 Classification of speaker recognition

III. EXTRACTION OF SPEECH FEATURES

A speech signal has several features such as phonetic,

prosodic and acoustic etc. Selection of the required speech

features among several features is the important task. By

selecting more informative speech features will help to

improve system performance. Selection of less useful or not

useful features for a particular task is unfavorable to the

system performance. Hence, examining the new speech

features for a specific task has been always an important and

difficult task [27]. Speech signals do not only carry language

information but it also includes key paralinguistic

components, prosody. Speech features stated as prosodic

features define prosody of a speech while the features which

are not used to describe prosody are called acoustic features

of speech. Fundamental frequency is the main component of

a speech signal which is further explained as:

Fundamental Frequency features: Fundamental

frequency (F0) features are useful for tonal languages. Tones

are related with dynamics of the fundamental frequency. To

extract F0 there are many methods used. One common

method is through autocorrelation i.e. autocorrelation of the

signal within a frame. In this computation, second highest

peak of autocorrelation is represented as fundamental

frequency of speech signal. For better accuracy or to make

system more robust against noise another technique is

required. It is based on observation such as tracking the peak

of the autocorrelation across frames or normalizing the

autocorrelation according to the analysis window [15] [28-

29] [30].

Measurement unit of F0 is Hertz (Hz) i.e. number of

periods within one second. Fundamental frequency period

can be further divided as jitter and shimmer.

Jitter: It is frequency stability in terms of equality of

period‟s duration. It is computed as average absolute

difference between consecutive periods (divided by the

average period) [31-32].

Shimmer: It is the measurement for amplitude stability of F0

periods. It is computed as Average absolute difference

between the amplitudes of sequential periods (divided by the

average amplitude) [31-32].

Fig. 4 Process of speaker verification system

Both jitter and shimmer have their observation based

thresholds and basically used in speech pathology research.

From the fundamental frequency and glottal cycle point of

view, there are many phonation types voice such as the

following [15]:

Normal Voice: Vocal cords are in their natural mode [15]

[16].

Creaky Voice, Vocal Fry: Creaky phonation is

characteristically related with aperiodic glottal pulses. In

such type of voice, degree of aperiodicity in the glottal

source is quantified by measurement of the jitter. During

creaky phonation, jitter values are higher than other

phonation types [14-16].

Falsetto voice: in such type of voice production vocal folds

are stretched longitudinally (thin). Therefore vibrating mass

is smaller hence tone is higher [15].

Breathy voice: It is noticeable as compound phonation type.

Such type of voice production has vocal fold vibration which

is inefficient due to incomplete closure of the glottis [15].

F0 rises when the vowel follows an unvoiced volatile

and decrease when it follows a voiced explosive.

IV. APPLICATION OF SPEAKER RECOGNITION In the last few years use of biometric system has become a

reality. There are lots of commercial as well as personal

applications where biometric is used for security purpose.

Speaker verification has gained a huge acceptance in both

government and financial sectors for secure authentication

[5] [33-34]. Australian Government organization Centrelink,

use speaker verification for authentication of recipients using

telephone transactions [35]. Possible applications of speaker

recognition are forensic investigation, telephone banking,

access control, user authentication etc. [36].

Speaker recognition have more potential than other

biometrics such as face recognition, finger prints, and retina

scans. The main advantage of speaker recognition over other

biometric is low cost, high acceptance and non-invasive

character of speech acquisition. To develop a speaker

recognition system, expensive equipment as well as direct

participation of speakers is not required. Speaker recognition

have potential to eliminate the need of carrying debit card,

credit card, remembering password for bank account or any

other security locks and many other online services [6] [37-

38]. With the continuous improvement in reliability of

speaker recognition technology, its usability has increased.

Now days, use of speaker recognition has become a

commercial reality and part of consumer‟s everyday life [34]

[39].

The performance of speaker recognition system is

vulnerable to change in speaker characteristics such as age,

health problems, speaking environment etc. Another

disadvantage is that it is possible to play a recorded voice

instead of the actual voice of a speaker [2] [34-35].

Access Control: Controlling access to computer

networks

Transaction Verification: For telephone banking and

account access control

Law Enforcement: Used in home parole monitoring and

residential call observing

Speech Data Organization: Used for voice mail. E.g.

speech skimming or audio mining applications.

Personalization: Device customization, store and fetch

personal information based on user verification for multi

user device [33].

All the above mentioned applications require robust

speaker recognition techniques. The requirement of

robustness in case of speaker recognition system can be

explained with the help of an example. In telephone based

services a user speaks in many circumstances (in noisy

environment or street), use different communication medium

(telephone or mobile), differs the distance of microphone etc.

Therefore robustness is the key factor for deciding the

success of speaker recognition system. In the area, in 1983,

the first international patent was filed by Michele

Cavazza, Alberto Ciaramella. This invention relates to

analysis of speech characteristics of speakers, in particular,

to a device for a speaker's verification [40].

Barclays Wealth and Investment Management was the

first organization in the world to deploy passive voice

security services. The basic aim behind using voice based

security was transforming the customer service experience.

By using this technique customers are automatically verified

as they speak with a service center executive [41]

In August 2014 GoVivace, developed speaker

identification system by using voice biometric technology.

This technology can be used for rapid voice sample matching

with thousands or millions of voice recordings. The purpose

of implementing this technology is to identify callers in

enterprise contact canter settings where security is a major

concern. GoVivace's SI technology is also available as an

engine. Application Programming Interfaces (APIs) to use

the software as a service [58]. In the UK, HSBC is

developing voice recognition and touch ID services for 15

million customers by the summer in a big step towards

biometric banking [42].

The popularity of voice biometric has risen more in past

few years. According to Opus research, more than a half

billion voiceprint will be in record, alone by 2020. People

found more comfortable with biometric authentication [3].

V. CONCLUSION

This paper provides concise definition and discussion about

speaker recognition technology. In addition, basic concepts

of automatic speaker recognition systems, modeling

technique etc. has been discussed. Speaker recognition is

method of designing a system for identity of an individual

through their voices. Speaker recognition has a significant

prospective as it is appropriate biometric technique for

security. The speaker recognition task is normally achieved

by acquiring speech signal, feature extraction, modeling

speech features for speaker, pattern matching and obtaining

match score.

REFERENCES

[1] B. S. Atal, “Automatic Recognition of Speakers from their Voices”,

Proc. IEEE, vol. 64(4), pp. 460-475, 1976.

[2] Q. Jin, “Robust Speaker Recognition”, PhD Thesis, Language

Technologies Institute School of Computer Science Carnegie Mellon

University, Pittsburgh, pp. 23-177, 2007.

[3] A. Rajsekhar G., “Real Time Speaker Recognition using MFCC and

VQ”, PhD Thesis, Department of Electronics & Communication

Engineering, National Institute of Technology Rourkela, pp. 9-71,

2008.

[4] J. Luettin, “Visual Speech and Speaker Recognition”, PhD Thesis,

Department of Computer Science University of Sheffield, pp. 16-156,

May 1997.

[5] T. Kinnunen, “Spectral Features for Automatic Text-Independent

Speaker Recognition”, Licentiate‟s Thesis, University of Joensuu,

Department of Computer Science Joensuu, Finland, pp. pp. 1-151,

December 21, 2003.

[6] M. R. Srikanth, “Speaker Verification and Keyword Spotting Systems

for Forensic Applications”, PhD Thesis, Department of Computer

Science and Engineering Indian Institute of Technology Madras, pp.

1-135, Dec. 2013.

[7] U. Sandouk, “Speaker Recognition Speaker Diarization and

Identification”, PhD Thesis, University of Manchester, School of

Computer Science, pp. 14-101, 2012.

[8] M. Savvides, “Introduction to Biometric Technologies and

Applications”, ECE & CyLab, Carnegie Mellon University.

http://www.biometricscatalog.org/ biometrics/biometrics_101.pdf or

biometrics_101.pdf

[9] M. El. Ayadi, Abdel-Karim S.O. Hassan, Ahmed Abdel-Naby, Omar

A. Elgendy, “Text-independent Speaker Identification using Robust

Statistics Estimation”, Speech Communication vol-92, pp. 52–63,

2017.

[10] S. Sarkar, Sreenivasa, Rao, K., “Stochastic Feature Compensation

Methods for Speaker Verification in Noisy Environments” Appl. Soft

Comput. 19, pp. 198–214, 2014.

[11] G. Doddington, Liggett, W., Martin, A., Przybocki, M., and Reynolds,

D, “Sheeps, Goats, Lambs and Wolves: A Statistical Analysis of

Speaker Performance in the NIST 1998 Speaker Recognition

Evaluation”, In Proc. Int. Conf. on Spoken Language Processing

(ICSLP 1998) (Sydney, Australia), 1998.

[12] S. Furui, “Speaker-Dependent-Feature Extraction, Recognition and

Processing Techniques”, ESCA Workshop on Speaker

Characterization in Speech Technology Edinburgh, Scotland, UK, pp.

10-27, June 26-28, 1990.

[13] L. P. Cordella, P. Foggia, C. Sansone, M. Vento, “A Real-Time Text-

Independent Speaker Identification System”, Proceedings of the

ICIAP, pp. 632, 2003.

[14] B. Richard Wildermoth, “Text-Independent Speaker Recognition

Using Source Based Features”, PhD Thesis, Griffith University

Australia, pp. 1-101, Jan. 2001.

[15] M. Breen, L. C. Dilley, J. Kraemer, and E. Gibson, “Inter-Transcriber

Reliability for Two Systems of Prosodic Annotation: ToBI (Tones and

Break Indices) and RaP (Rhythm and Pitch)", Corpus linguist. ling.,

vol. 8 (2), pp. 277-312, 2012.

[16] A. Stolcke, “Higher‐Level Features in Speaker Recognition”, Winter

School on Speech and Audio Processing, IIT Kanpur, Jan. 2009.

[17] N. Dehak, D. Pierre, and K. Patrick, “Modeling Prosodic Features

with Joint Factor Analysis for Speaker Verification”, IEEE

Transactions on Audio, Speech, and Language Processing, vol. 15 (7),

pp. 2095-2103, Sept. 2007.

[18] Z. Huang, L. Chen and M. Harper, “An Open Source Prosodic Feature

Extraction Tool”, School of Electrical and Computer Engineering

Purdue University West Lafayette, pp. 2116-2121, 2006.

[19] T. Kinnunen and Haizhou Li, “An Overview of Text-Independent

Speaker Recognition: From Features to Supervectors”, Speech

Communication vol.52, pp.12–40, 2010.

[20] N. Singh and Khan R. A., “Underlying of Text Independent Speaker

Recognition”, in IEEE Conference (ID: 37465) (10th INDIACom

2016 International Conference on Computing for Sustainable Global

Development), held on 16th -18th March, 2016 at BVICAM, New

Delhi, pp. 11-15, 2016.

[21] D. Sierra Rodriguez, “Text-Independent Speaker Identification”, PhD

Thesis, AGH University of Science and Technology Krakow, Faculty

of Electrical Engineering, Automatics, Computer Science and

Electronics, pp. 1-121, 2008.

[22] M. Ghassemian and K. Strange, “Speaker Identification - Features,

Models and Robustness”, Technical University of Denmark, DTU

Informatics Kongens Lyngby, Denmark, pp. 1-118, 2009.

[23] D. A. Reynolds, “An Overview of Automatic Speaker Recognition

Technology”, MIT Lincoln Laboratory, Lexington, MA USA,

@2002IEEE, pp. 4072-4075, 2002.

[24] B. S. Atal, “Effectiveness of Linear Prediction Characteristics of the

Speech Wave for Automatic Speaker Identification and verification”,

Journal Acoustical Society of America, vol. 55, no. 6, pp. 1304-1312,

June 1974.

[25] "Speaker Identification". Archived from the original on August 15,

2014. Retrieved September 3, 2014.

[26] A. Majetniak, “Speaker Recognition using Universal Background

Model on YOHO Database”, Aalborg University, The Faculties of

Engineering, Science and Medicine Department of Electronic Systems,

pp. 1-61, May 31, 2011.

[27] B. Martínez-Gonzalez, M. Jose Pardo, D. Julian Echeverry-Correa,

Ruben San-Segundo, “Spatial Features Selection for Unsupervised

Speaker Segmentation and Clustering”, Expert Systems with

Applications, vol. 73, pp. 27–42, 2017.

[28] D. Talkin, “A Robust Algorithm for Pitch Tracking (RAPT)”, Speech

Coding and Synthesis, pp. 495-518, 1995.

[29] P. Boersma, “Accurate Short-term Analysis of the Fundamental

Frequency and the Harmonics-to-Noise Ratio of a Sampled Sound”,

In Proc. the Institute of Phonetic Sciences, 1993.

[30] C. Liu, P. Jyothi, H. Tang, V. Manohar, M. Hasegawa-Johnson, and S.

Khudanpur, “Adapting ASR for under-Resourced Languages using

Mismatched Transcriptions”, In Proc. ICASSP, 2016.

[31] E. Vayrynen, “Emotion Recognition from Speech Using Prosodic

Features”, PhD Thesis, University of Oulu Graduate School,

University of Oulu, Faculty of Information Technology and Electrical

Engineering, Department of Computer Science and Engineering,

Infotech Oulu, pp. 1-92, 2014.

[32] N. Singh, A. Agrawal and Khan R. A., “Automatic Speaker

Recognition: Current Approaches and Progress in Last Six Decades”,

Global Journal of Enterprise Information System. Vol. 9, Issue-3,

July-September, pp. 38-45, ISSN: 0975-1432, 2017.

[33] D. A. Reynolds, “Automatic Speaker Recognition: Current

Approaches and Future Trends” MIT Lincoln Laboratory, Lexington,

MA USA, pp. 1-6, 2001.

[34] S. Memon, “Automatic Speaker Recognition: Modeling, Feature

Extraction and Effects of Clinical Environment”, PhD Thesis, School

of Electrical and Computer Engineering Science, Engineering and

Technology Portfolio RMIT University, pp. 1-242, June 2010.

[35] R. Summerfield, T. Dunstone, C. Summerfield, “Speaker Verification

in a Multi-Vendor Environment”,

www.w3.org/2008/08/siv/Papers/Centrelink/w3c-sv_multivendor.pdf

[36] R. Nakatsu, J. Nicholson, and N. Tosa, “Emotion Recognition and its

Application to Computer agents with Spontaneous Interactive

Capabilities”, Knowledge based systems, vol. 13, pp. 497–504, 2000.

[37] G. Gravier and G. Chollet, “Comparison of Normalization Techniques

for Speaker Verification”, In Proc. on Speaker Recognition and its

Commercial and Forensic Applications (RLA2C), pp. 97-100, 1998.

[38] N. Singh, Khan R. A. and Raj shree, “Applications of Speaker

Recognition” Science Direct, Procedia Engineering, vol-38, pp.

3122-3126, 2012.

[39] E. Shriberg and A. Stolcke, “Direct Modeling of Prosody: An

Overview of Applications in Automatic Speech Processing”, In

International Conference on Speech Prosody, 2004.

[40] A. Michele Cavazza and C. Alberto, "Device for Speaker's

Verification”,

" http://www.google.com/patents/US4752958?hl=it&cl=en

[41] International Banking (December 27, 2013). "Voice Biometric

Technology in Banking | Barclays". Wealth.barclays.com.

Retrieved February 21, 2016.

[42] K. Julia, "HSBC Rolls out Voice and Touch ID Security for Bank

Customers | Business", The Guardian, Retrieved February 21, 2016.

Examine the aptness of strategies of

crowdsourcing in agile software methods

Himanshu Pandey

Research Scholar

MUIT, Lucknow.

hpandey010@gmail.com

Dr. Santosh Kumar

Associate Professor,

MUIT, Lucknow.

sant7783@hotmail.com

Dr. Manuj Darbari

Associate Professor,

BBDNITM, Lucknow.

manujuma@gmail.com

Abstract- Agile software development has been vigorous research

domain for more than a decennary. Nonetheless, forthright

application of agile methods in actual praxis is encumbered by

few confront, which require to be resolute both by industry and

academia. One of the extensive off spring in this respect solicitude

is the dearth of resources, assets and funds for creation of

application and development of agile methodology and

techniques. The crowdsourcing methods in the area or field of

software development like in Top Coder and Apple App have

shown promising and feasible solutions to various problems. Use

of crowd methodology in the field of software development is

estimated to take its place along with existing methodologies or

techniques like service-oriented computing, agile. In this study,

we put forward a hypothetical and inquisitive framework of

crowdsourcing strategies to anatomize its existence and

emergence in agile software development. The paper has been

put forward with the concept of crowdsourcing cognizance into

the current agile methods along with apprehension the strategies

of crowdsourcing approach in agile software development. To

scrutinize the applicapability and existence of crowdsourcing

strategies like Crowd creation, Crowd Democracy, Crowd

Reviews, Crowd Wisdom and Crowd funding, we exploited both

quantitative and qualitative research methods. We also describe

the basal principles of crowdsourcing strategies in agile

development process.

Keywords: Crowdsourcing, Crowd funding, Crowd creation,

Crowd reviews, Crowd democracy, Crowd Wisdom

I. Introduction

Software engineering is limited to small association of

developers, but is gradually widespread in, communities,

institutes and organizations involving large group of peoples.

There is an increase in globalization with a stand-point on

coordinated practices and infrastructure. Only evolving

method getting work completed is crowdsourcing, a sourcing

technique that discovered in the 1990s. Determined by Web

2.0 technologies, organizations, with an Internet connection

can form a workforce consisting of anyone. Customers can

publish blocks of tasks, or work, on a crowdsourcing platform,

where scrum or individual workers opt for those works that are

according to their skills and capabilities.

The commonly used methods for solving these problems

involve carrying out surveys, formation of committees and

forums, and hiring consultants. A new and efficient way of

solving these problems can be—crowdsourcing.

Firstly, all social networks like facebook, Twitter, and Skype

have changed the very meaning of its presence. The web

medium permits for collective intellectual abilities of a

population in methods. Due to time and place limitations,

face-to-face communication cannot possible in every time.

Our case study says that the crowdsourcing model is a

prosperous, attainable web-based, distributed problem solving

technique and production method for small business. It is an

effective model for enabling software development

procedures. Crowdsourcing has matured as a substantial

feature for development of any agile products, because it

permits the candidness and sharing of resources and

knowledge collected by communities. Crowdsourcing is a way

by which Software developers, designers, testers and others

not only develop framework for any software or application,

but also manages the collective resources of various

contributors and customizing products according to their goal

needs. Completing or solving a full project or software related

problem can be a difficult job for few people in the

organizations. With crowdsourcing, even non-professional

enthusiasts have the chance to participate in the development

of new types of agile software applications.

II. Literature Survey

Based on the literature review we analyzed various work and

task consort to crowdsourcing which are playing major role in

rapid and flawless software development process. The paper

by J.Bishop[1], investigate a study on Big Data from a project

called “QPress”. It was implementing by an organization, lie

on dependent working with inter-professionalism. This study

shows the value of the Cloud to distance working, like

teleworking means, individuality of the organization and how

can helpers associate to it, including what component play

major role on research worker motivation. This case study

examines that the organizational structure of the company

analyses, where many software companies behave different

works could be investigate as better exercise in assisting inter-

professional atmosphere. Monika Skaržauskaitė[2], undertakes

the benefit of crowdsourcing in educational activities. Focus

of this paper is to take deep investigation at how educational

organizations are employing crowdsourcing as component of

their act along with intimate how the praxis of crowdsourcing

may divide to another educational activity. Upendra Chintala,

Shrikanth, Anurag Dwarakanath, Gurdeep Virdi, Alex Kass+,

Anitha Chandran[3], analyses a numerous software

development methodology, gives main confront for the

assiduity of Crowdsourcing to the development of enterprise

application. This research showed a concept to methodically

break down the whole business application into small works

so that the works can be implemented autonomously in

equidistant by the crowd. This research endorses automated

testing and integration. Thierry Buecheler, Jan HerikSeig,

Rudolf M. Fuchslin and Rolf Pfeifer [5], clearly describe how

advantages of crowd, enjoyed by nonprofit “Research Value

Chain”. Research content explains a) how crowdsourcing can

be devoted to fundamental science and b) the impact of

conclusions of Artificial Intelligence research on activeness of

crowdsourcing. Thierry merged different research series of

Conclusions and results like, Versatile or complicated

networks. The research and works of Eddy Maddalena,

Vincenzo Della Mea, Stefano Mizzaro[6] , focuses the

probabilities that if a task applied on crowdsourcing platforms

is suitable two mobile devices. Eddy first aim is to investigate

the compatibility of preexisting crowdsourcing platforms on

mobile devices, along with what sort of the application of

Crowds express the benefits of crowdsourcing in educational

activities. In this paper [7], “The application of Crowdsourcing

in educational activities”, enlighten the existence of Crowd

Sourcing in educational activities. With the time passes, the

roar in Information and Communication Technologies by dint

of Internet and time limit of a wide area of alternatives for

organizations to attain their aim. The main concept of this

research is to give up deep elaboration of achievement of

crowdsourcing as part of their activities with explanation of

how implementation of outsourcing would be soon spread to

other educational activities. Andre VnHoek's [8],

“Crowdsourcing in Software Engineering, Models,

Motivations and Challenges” is about the possibility of crowd

in software engineering and advantages of crowdsourcing are

applicable in software or not. Crowdsourcing has begun

developing a place in software development area. The paper

[9],“Reactive Crowdsourcing”, by Alessandro Bozzon, Macro

Brambilla, StafanoCeri, Andrea Mauri proposers the aspects

of Crowd Sourcing which delivers find level, ex-controls and

powerful. We extract the initially design, crowdsourcing

application as commingle of basal task types. Further, we

reform this hi-tech feature into the specification of a reactive

execution environment; endorse planning, execution and

termination of a task as well analysis of performer. Controls

are established as live phenomenon in data structures which

are gathered from the specification and design of application.

These phenomenon can be replaced or altered as per

requirement, assures highest competence with the lowest

attempt. The paper by Kathryn T. Stolee, Sebastian

Elabaum[10], examine the benefit of crowdsourcing as a

procedure to discourse that confront by helping in subject

recruitment. Further, by help of his research we can expose

how study can be accomplished below infrastructure that

certify the existence to formulate a large base of uses and also

admit managing users while the research is running out. We

discuss the results and observations of this incident which

describes the efficiency and potential of crowdsourcing

software engineering studies. RajanVaish[11], has shown a

research direction which examines the suitability of expert

outsourcing by merging mentor with student crowd. This

procedure will permit mentors to alternatively used operators

such as split commerce, remove or add on project ideas, cord

on studying to daddy suit on project ideas. This research

process will contain numerous stages like paper-pencil

prototyping, brainstorming and development and user

evaluation for producing publishable results. This research

presents the existence of crowdsourcing research process with

operators along the research level, while solving the

constraints of resource and opportunity among mentor and

crowd. We are inspired by the work of Bernardo A.

Humberman[12], research of a large data set fetch from

YouTube, potential exhibited in crowdsourcing provides a

strong credence on attentiveness, composed by number of

downloads. Research paper by Yanni Han, Hanning Yuan, Jun

Hu[13], for a research perspective, propound a agile

development procedure of software ,based on service provider.

Along with, we investigate the service efficiency of software,

increased by the user-driven needs in network environment.

The agile development process is proposed based on users'

individual needs. This method can be represented as

demanded service resources and approach to the target slowly.

Tom Narock and Pascal Hitzler[14], for proving reliable

solution implementing search algorithms with crowdsourcing.

They observe Big Data within the geosciences and define

logical questions related to the merger of crowdsourcing. De

establishes a crowd Sourcing portal that permits people of the

jio science community to club their conference presentations

and funded grant descriptions to the Database used in those

projects. Rashmi Popli, Naresh chauhan[15], cast the spotlight

on the research work in Agile Software Development and

calculation in Agile. They favored a technique for the actual

cost, to avoid from the problems of current agile practices.

Related to the concepts of flexibility along with adaptability,

the agile methods, shown an innovative set of software, are

presently used as a gimmick to these recurrent problems and

make but a clear way for the future of development. Emal

Altameem[16], proposed several methods , in which Agile

concepts has been, proved to be influential in software

development. It also defines the advantages and the limitations

of Agile Technique. This research motivates developers, to

assume this technique, to develop software that proves out to

be a remedy for their changing requirements. Gaurav Kumar,

Pradeep Kumar Bhatia [17], resolves the significance of this

methodology, in terms of its quality within the cultural

framework. KudaNageswaraRao, G.KavitaNaidu, Praneeth

Chakka[18] has been executed with the factual intentions of

investigation and gain acuity. Into, the latest agile concepts

and Techniques, distinguishes between capability and

weaknesses of agile methods and numerous issues regarding

their applicability. WenjunWu, Wei-Tek. Tsai [19], examine

the data in detail, gathered on software crowdsourcing and

abbreviates major lessons, learned, then analyze two Software

crowdsourcing processes, containing artistic ways. Preeti Rai,

Saru Dhir[20],elucidate the influence and comparison of

several traditional techniques and a new methodology. Top

Coder and App Story processes. Concluding, they identify the

min-max nature among candidates as a crucial element of

design in software Crowd-Sourcing for software quality and

investigate the reasons for which software industries, shifted

from Traditional RE to Agile RE. Gaurav Kumar, Pradeep

Kumar Bhatia [21], recognize the fact agile methodology has a

remarkable impact over software development procedure

related to quality, within the methodical, organizational and

cultural framework. Kiran Jamaalam adakal, V. Rama Krishna

[22], emphasize on few confrontations with Agile->scrum and

gives vision to the user whether the Agile is WONDER

DRUG. Malik Hneif, Siew Hock Ow[23], showcase their

review over three Agile approaches (Extreme programming,

Agile Modeling, and scrum) distinguishes between them and

advice, when to adapt them. Gurleen Singh, Tamanna[24],

reviews various agile techniques like on their characteristics,

objectives, boons and banes of using agile methodology and

their unique characteristics.

III. Suitability of Crowd sourcing strategies in Agile

Methods

A. Crowd Creation

Crowd creation is the best usual constitute of crowdsourcing.

It is related to build up activities of software development

process, asking personage and business to clear up a specific

state of uncertainty, so that produce an adequate answer to the

particular problem. End product is the output of the crowd

creation, even if in intellectual capital or physical form

(software product) , that has a concrete value to remains. One

of the three forms (Cattle call, Warehousing Mechanism,

collective project) is pursuing by crowd creation.

a) Cattle Call

In the first form of crowd creation, to achieve the specific

need, company or individuals places a request for submission.

Instances of this are slogan, fan generated commercials or

design concepts. These have a clear cut task of a product

promotion or work completion. Individuals or scrum master

convey their complete design to the wishing entity in their

complete form. Selection process has followed for choosing

either a small group or single selection. This selection process

can be done through crowd voting or inside the company

professionals.

b) Warehousing Mechanism

As same to the cattle call, scrum master or entity places a

requirement on the web for any project or task in warehousing

mechanism. The specific content and submission do not

achieve by the objective of the site, that type of restriction

proposed by the requester. Software organization or software

professionals upload their innovative creation or product to the

base entities website to be seeable by the people at large.

Dissimilar cattle call, foe selection process, there is no criteria

followed by the company; those items are displayed that

achieve to guidelines or instructions. Customers have a

flexibility to choose either the developed software product or

the raw content itself. The base company allows the customer

selection to them in the sale form. After completion of the

transaction, revenues sale are divided into the software

company and the content provider at a fated rate. One another

limitation of the form is, possession authority of the image

may or may not with the begetter.

c) Communal project

The main contrast among other forms of crowd creation and

collective project is that, from the crowd are member of whole

project, and creates the end product. In the era of internet,

centers basically close to cerebral capital as adverse to

physical formation like programming code. Collective project

has two forms.

i. Cattle call of multistage- Contains all these

process:

• Forthcoming with the new concept

• Determine either it creates a attainable production

candidate

• Examining the completed design of the project

• Generating the final item is break down into

individual chunks

ii. Open source initiative- Generally, application of

that type of project drawn under the form of open

source initiatives or micro work. Commonly,

solutions have to adjust inside the framework of

the application.

B. Crowd Wisdom

Crowd Wisdom is the process that takes into consideration.

The collective opinion of a group of individuals, having their

own point of views than a single person. This theory states that

answers gathered and discussed with large groups that involve

estimation, reasoning, worldly affairs are usually better.

Crowd Wisdom aims at less, but better output and aggregates

individual knowledge. It needs explicit command and

motivation.

The Key conditions for Crowd Wisdom are:

1. Decentralization->Cognition-> Co-operation

2. Diversity of opinion ->Co-ordination

a) Retrospectives

Retrospectives, in the perspective of crowdsourced agile

software development, are the conversations that are held after

frequent intervals or iterations. During these conversations, the

team reflects on what happened in the previous iterations and

identifies the areas where more improvement is desired or

needed, and takes or plan future actions accordingly. During

these conversations an environment of trust and honesty are

necessary needed so that each member can share their

thoughts without hesitating .This also boosts their confidence

levels.

C. Crowd Democracy

Crowd Democracy is also used in policy making in project; it

also considers outer professionals opinion and gathers

information. The US Federal government uses Crowd

Sourcing for several years as a way of citizen participation in

national open government strategy. Transparency and

collaboration being at its core.

Crowd sourcing is also used in every phase of agile software

development. Even though, Crowd Sourcing is a new and

emerging technology it has been used in various projects

initiated by the software companies all around the globe.

With the help and assistance of Crowd Democracy companies

or individuals can get direct feedback from its requester or

customers. It has also been used in many Non- Profit

Organizations. After developing the project under crowd

sourcing method, crowd democracy plays a major role to

insure their quality attributes.

A project was carried out to test crowd sourcing is public

participation for transit planning in salt take from 2008 to

2009 founded by USA’s federal transit administration grant.

Sometimes, it is difficult to gather demographic data of crowd.

Both intrinsic and extrinsic motivation cause people to worse

for or contribute to a crowded task.

D. Crowd Reviews

Main perspective of crowd reviews is taking task requirements

or reviews from the crowd for the software development.G2-

Crowd is a peer to peer, business solution review platform.

The company was established in 2012 and it provides real time

and unbiased reviews based on user rating and other relevant

data to help you assess what is best for business.

There are few problems that always take place with

retrospectives some problems are like team members are not

able to understand the full value of a retrospective meeting and

even some are unwilling to participate in it questioning its

importance.

Now, here we have a few problems along with their

explanation and solutions. They are as follows:

• Thinking of one’s team to be a perfect for

retrospectives. It is almost impossible that any team

is so perfect that they need no improvements. Even if

this happens then it means that the team is not

progressing. Therefore, no team is too good.

• Sometimes members find the meetings boring and

monotonous. So , it was never mentioned anywhere

that they should be exciting after all it is a meeting

and it should be carried out following all rules and

regulation. And the decorum should always be

maintained.

• There are even people who say that they don’t like

retrospectives. Now, it is not possible to like all

aspects of jobs but the sake of professionalism

participate in each and every mandatory meeting or

gathering or discussions.

E. Crowd funding

In a software development field every professional is trying to

get money for their concept or innovative ideas, in the

erstwhile, venture capital funds or angel investors were

options for getting large sums of cash. These days, crowd

funding has become a pragmatic resource for software

developers or entrepreneurs who require assets but do not wish

to forfeit investment or go throughout the arduous procedure

of being interrogate by financier.

Crowdfunding is a novel method for entrepreneur or software

developer to elevate capital by tapping the masses rather than

a small pool of financier. Some instances might have heard of

a few, like kickstarter helps to find the resources and support

they wish to form their innovative ideas an actuality, to help

bring creative projects of life. According to our analysis,

“kickstarter is the best way to connect with people who can

truly help you.

Crowd funding works by permitting in essence anyone to

compose relatively small assistance to a project in interchange

for tangible goods, rather than having a small group of

financier or investors, risk big chunks of funds. Contribution

funds variegate from campaign to campaign but generally

range from dollar 1 to dollar 10,000. Crowdsourcing sites

basically primitively ill fame in the creative and innovative

communities of groups.

IV. FUTURE SCOPE

Crowdsourcing strategies will be segmented i.e., for each

subtask, professional organizations will evolve and occupy the

market which may or may not create a monopoly.

Crowdsourcing strategies (Crowd funding, Crowd creation,

Crowd reviews, Crowd democracy, Crowd Wisdom) will be

applied to different fields of software development and

specification will be strict and bounded. The input and output

of the crowdsourced task will be strictly defined and

constrained which in turn will result in overall quality

improvement of the output. More and more developers will be

attracted towards this method as it is not bounded by any

factors like deadline pressure and will enable them to work as

per their interest and emerge with quality solutions.

REFERENCES

[1] J. Bishop, “Supporting crowd-funded agile software

development projects using contingent working: Exploring

Big Data from participatory design documentation and

interviews”, Int'l Conf. Information and Knowledge

Engineering |IKE’15.

[2] Monika Skaržauskaitė, “The application of crowdsourcing

in educational activities”, ISSN 2029-7564 (online)

SOCIALINĖS TECHNOLOGIJOS SOCIAL

TECHNOLOGIES 2012, 2(1), p. 67–76.

[3] Anurag Dwarakanath, Upendra Chintala, Shrikanth N. C.,

Gurdeep Virdi, Alex Kass+, Anitha Chandran, “ CrowdBuild:

A Methodology for Enterprise Software Development using

Crowdsourcing”,

http://digitalworkforce.accenture.com/axpense/.

[4] Kenneth Benoit, Drew Conway, Benjamin E. Lauderdale,

“Crowd sourced text analysis: Reproducible and agile

production of political data”, Slava Mikhaylov University

College London .June, 2015.

[5] Thierry Buecheler, Jan HenrikSieg, Rudolf M. Füchslin1

and Rolf Pfeifer, “Crowdsourcing, Open Innovation and

Collective Intelligence in the Scientific Method: A Research

Agenda and Operational Framework”, Artificial Intelligence

Laboratory, Department of Informatics, University of Zurich.

[6] Vincenzo Della Mea, Eddy Maddalena, Stefano Mizzaro,

“Crowdsourcing to Mobile Users: A Study of the Role of

Platforms and Tasks”, DBCrowd 2013: First VLDB Workshop

on Databases and Crowdsourcing.

[7] Monika Skaržauskaitė “The application of crowd sourcing

in educational activities”, ISSN 2029-7564 (online)

SOCIALINĖS TECHNOLOGIJOS SOCIAL

TECHNOLOGIES 2012, 2(1), p. 67–76.

[8] André van der Hoek, “Crowdsourcing in Software

Engineering Models, Motivations, and Challenges”, focus:

The Future of Software engineering.

[9] Alessandro Bozzon. Marco Brambilla, Stefano Ceri,

Andrea Mauri, “Reactive Crowdsourcing”, Dipartimento di

Elettronica, Informazione e Bioingegneria – Politecnico di

Milano Piazza Leonardo da Vinci, 32 – 20133 Milano, Italy

[10] Kathryn T. Stolee, Sebastian Elbaum, “Exploring the Use

of Crowdsourcing to Support Empirical Studies in Software

Engineering”, Department of Computer Science and

Engineering University of Nebraska – Lincoln, NE, U.S.A.

[11] RajanVaish, “Crowdsourcing the Research Process”,

University of California at Santa Cruz1156 High Street, Santa

Cruz, California 95064, USA.

[12] Bernardo A. Huberman “Crowdsourcing, attention and

productivity”, Social Computing Lab, HP Laboratories, Palo

Alto, CA, USA.

[13] Hanning Yuan, Yanni Han, Jun Hu, “Research on Agile

Development Methodology of Service-Oriented Personalized

Software” published by IEEE xplore ,2008 International

Conference on Computer Science and Software Engineering.

[14] Tom Narock and Pascal Hitzler “Crowdsourcing

Semantics for Big Data in Geoscience Applications”,

Semantics for Big Data AAAI Technical Report FS-13-04.

[15] Rashmi Popli, Naresh Chauhan, “Cost and effort

estimation in agile software development”, published by IEEE

2014 International Conference on Reliability Optimization and

Information Technology (ICROIT).

[16] EmanA. Altameem, “Impact of Agile Methodology on

Software Development” published by Computer and

Information Science; Vol. 8, No. 2; 2015 ISSN 1913-8989 E-

ISSN 1913-8997 Published by Canadian Center of Science

and Education 9.

[17] Gaurav Kumar, Pradeep Kumar Bhatia,“Impact of Agile

Methodology on Software Development Process” published

by International Journal of Computer Technology and

Electronics Engineering (IJCTEE) Volume 2, Issue 4, August

2012.

[18] KudaNageswaraRao, G. Kavita Naidu, PraneethChakka,

“A Study of the Agile Software Development Methods,

Applicability and Implications in Industry”, published by

International Journal of Software Engineering and Its

Applications Vol. 5 No. 2, April, 2011.

[19] Wenjun Wu, Wei-Tek Tsai, “Creative software

crowdsourcing: from components and algorithm development

to project concept formations”, published by Int. J. Creative

Computing, Vol. 1, No. 1, 2013.

[20] PreetiRai, SaruDhir, “Impact of Different Methodologies

in Software Development Process” published by ,

International Journal of Computer Science and Information

Technologies, Vol. 5 (2) , 2014.

[21] Gaurav Kumar, Pradeep Kumar Bhatia,“Impact of Agile

Methodology on Software Development Process” published

by, International Journal of Computer Technology and

Electronics Engineering (IJCTEE) Volume 2, Issue 4, August

2012.

[22] Kiran Jammalamadaka1, V Rama Krishna, “Agile

Software Development and Challenges” IJRET: International

Journal of Research in Engineering and Technology. ISSN:

2319-1163, ISSN: 2321-7308 Volume: 02 Issue: 08 | Aug-

2013.

[23] Malik Hneif, Siew Hock Ow, “Review of Agile

Methodologies in software development” International Journal

of Research and Reviews in Applied Sciences ISSN: 2076-

734X, EISSN: 2076-7366 Volume 1, Issue 1(October 2009).

[24] Gurleen Singh, Tamanna ,“An Agile Methodology Based

Model for Software development”, published by, International

Journal of Advanced Research in Computer Science and

Software Engineering Research Paper, Volume 4, Issue 6,

June 2014.

IMPLEMENTATION OF ARTIFICIAL INTELLIGENCE AND

NEURAL NETWORK LOGIC IN CLOUD COMPUTATION

Shailendra Kumar Rawat1, S. K.Singh

2, Amar Nath Singh

3,

1,3Bengal College of Engineering & Technology, Durgapur 2Amity Institute of Information

Technology, Lucknow

Email: rawat.nokia@gmail.com, sksingh1@amity.edu, amarnathsingh.2k@gmail.com

ABSTRACT

Cloud computation is the one of the most recent and

modern technique, which basically aimed to

maximize the benefit of distributed resources and

aggregate them to achieve higher throughput to solve

large scale computation problems. In this technology,

the customers needn’t require to buy the product for

their use rather they can use to process the task as per

the use basis. It is the computation technique which is

considered as the fastest computation because of its

architecture. As we know that due to the advance in

problem prospect, the complexity of processing the

task is getting more so the computation logic also

need to be enhanced accordingly. In this paper we

introduce the cloud computing technique using

generic approach of algorithm, Back propagation

technique of artificial Intelligence and artificial

neural networks, and then reviewed the literature of

cloud job scheduling. We here propose to

implementing artificial neural networks which is a

component of Artificial Intelligence to optimize the

job sending sequence for processing which is also

called as scheduling, results in cloud as it can able to

find the find new set of process classifications.

Keywords:

Cloud Computation, Job Sequencing, Artificial

Intelligence, Artificial Neural Networks, Back

propagation technique, generic algorithm.

1. Introduction As we know that in traditional days the computation

process was very time consuming and hectic due to

because of its job sequencing. But due to the

presence of cloud concept where the resources can be

used as per the requirement basis, the task became

more and more sophisticated to get processed. In the

cloud structure the arrangement of the resources can

be of various types , but here we have taken the

arrangement as showed in Figure 1. In traditional

days we need to access the resources in traditional

ways but due to the implementation of Cloud

technique, It makes the complexity of the problem

minimum by resizing the problem and shifts up the

IT infrastructure to such a great extent so that the

computation process don’t require the interpretation

of third party logic[1, 2, 9]. The process of

computing in cloud became more and more efficient

and economical and can be outsourced to another

party as and when required by the user. So by doing

this we reduces the complexities of processing of the

task and simultaneously we increase the job

sequencing to a great extent [3][4].

2. Cloud Computing Deployment:

As we know that, the cloud is a open wide access

prospect, so the computing of the task or accessing of

the task for processing in the cloud environment

became more sophisticated due to its characteristics

which is commonly known as the, customization of

task, portability of task fetching, availability of the

desired resources on users demand and isolation of

the task as and when required [6, 29]. It is also

important that, the computability shouldn’t get

effected by increase in terms of task ratio [7, 23]. In

modern technology of era, most of the IT Companies

and other sectors are using the cloud, because do not

need to invest a huge amount of budget in new

infrastructure and training to the employee to use the

cloud and its operation. Hence, by using cloud

computing, the Small and Medium scale category

Businesses (SMB) can able to access the best

applications for their purpose and resources for their

requirement at very low cost [8]. On the other hand

if we try to emphasize the security of the data then

cloud provides a better security other then the

traditional approach of security in terms of data

prospects for their use [8]-[10].

Whenever any companies try to implement the cloud

network then, the major task is its deployment.

Normally a cloud is consisting of various structures,

out of which the basic classification aspect it is of 4

types, such as:

1. public cloud which is a wide application

aspect,

2. private cloud used for personalize purpose,

3. hybrid cloud which is a combination of

public and private cloud and,

4. Community cloud which a special purpose.

In the public cloud as it is open to all, so if a person

wants then he may access it using any web browsers.

Thus by implementing this mechanism will decrease

the operation costs for processing the task. But

security point of aspect, they are less secure, because

we know that all the software which is used now a

day are much typical and data present on this model

are more exposure to outside world and a hacker can

easily hack the data or corrupt the data [11].

In the private cloud which is also a personal purpose

in terms of their use, and can be best for local area or

limited area use. This private model of cloud is

similar to the Intranet. The internet is just a common

gateway or a medium for which we are using it to

transfer the data. The main biggest advantage of

private cloud is that it is very easy to operate and the

data is very secure in this case. The operational cost

and maintenance of the infrastructure can be

upgraded as and when requires at low cost [4].

Similarly in case of hybrid model, as it is a

combination of private and public cloud so, in this

case, a private cloud is linked to one or more external

cloud services which in terns extends as public. [4]

[11].

To make the cloud computing system effective we

need a medium such as internet where there are front-

end and back-end. The front-end is mostly

responsible to connect with cloud and provides the

platform for the user to access the cloud. On the other

hand, the back-end part is the platform through which

the task result are going to processed and the results

are stored. Figure 2 shows the detailed approach of

the cloud structure using which we can easily used

the cloud as per our purpose and the mode of

operation. [6].

Figure 1. Cloud computing paradigm.

It provides the computing environment through the

platform by using which user can easily process the

task and can able to get the resource service as and

when required. It is a common platform for both the

front end as well as the back end [3][5].

Figure-2

3. Genetic Algorithm and Artificial

Neural Networks A generic Algorithms which is also some times

referred as Genetic algorithm is basically a part of

heuristic approach, using which we can find the

optimal natural selection and natural genetics. Using

this approach we can find the similar estimated

solutions to difficult problems, see Figure 3 [19]

[20]. The main principle behind the genetic algorithm

is to identify the problem scenario and conjugate

them to identify the solution. This algorithm was first

introduced in 1962 by Holland [21,22,25]. The

purpose of using the genetic algorithm here is to

identify the problem statement of the cloud in the

structure. As we know that when ever the task is

subject to process through the front end, then the job

sequencing became a typical problem. So to identify

the common ability of the problem is done using this

approach [33, 26,27]. In general way of its use , it

uses the concept of ANN model. The ANN is a

component of Artificial Intelligence, which is used to

identify the sensing of problem statement and its

related components. The ANN consist of three main

aspects, such as:

1. Network topology,

2. Transfer function, and

3. Training algorithm.

The ANN output accuracy is directly related with the

number of layers presents in the network [35]. Hence

depending on the parameters that used to train the

system. Networks performance is affected by number

of layers, ie if we increase the number of layers then

ultimately the performance of the system gets

decreases[34, 28,29]. So it is always suggested that,

to minimize the number of nodes layers and by doing

this we can enhance the performance of the system.

[26].

Figure 3. Simple genetic algorithm operation [24].

Figure 4. Basic structure of an artificial neural network

(ANN) [28].

4. Cloud Computing Job Scheduling

In every computational logic the process for

providing the task in the unit by the resource is very

important. Because it decides which task has to

processes at what sequence[30, 31]. Hence we

ultimately use the scheduling algorithms. In cloud

computing, the most important aspect is its

architecture. As we have already discussed abut the

various types of cloud category [20]. Out of them the

most primitive is their job sequencing and their

software and hardware configuration based on which

the entire job processing depends [29, 32].

6. Discussion

We know that, the cloud computing is the most

recent and modern trend of computing. Besides of

this there are also some other AI applications logic by

using which we can enhance the computation logic.

On the other hand the Neural Networks which is also

an another aspects of AI logic, provides a better

strategy for the cloud for its arrangement and its

mode of operation. As we know that once the cloud is

ready to use, then the job sequencing and identifying

the common aspect of job for the problem is typical.

Hence we must have to use the Genetic algorithm

along with ANN in the cloud to identify the aspect

and resolve the issue.

7. Conclusions and Future Works

As we know that the cloud computing the most recent

technology which are used widely for the computing

the complex problems. It is due to because of its

completely different architecture in term of its job

fetching and storing during the processing. It uses its

AI logic along with the ANN concept which in turns

uses the generic algorithm to identify the problem

scenario and enhance the job processing.

We also know that cloud uses the internet as its

medium to exchange the data , so we somehow think

for the security of data also. Although the cloud uses

its own security system but still we need to make

smart enough to prevent our data from the hackers

and external threats. It can be achieved by fusing the

AI logic in the modern software, so that it can be

identify during the processing itself.

About the Author:

Asst. Prof Shailendra Kumar Rawat is presently

pursuing his PhD in Computer Science stream at

Amity University, Lucknow.

Prof S. K. Singh is presently working as Reader in

Amity Institute of Information

Technology,

Lucknow. Prof. Amar Nath Singh is presently working at

Bengal College of Engineering, Durgapur as a reader

in the department of Computer Science and

Engineering. His research area is Underground Mines

and Surface mining using Artificial Intelligence,

Fuzzy Logic. His research area includes cloud

computing, WSN, Machine Learning and Data

Science. He has produced more than 60 M.Tech

scholars till date.

References

[1] Xu, X. (2012) From Cloud Computing to Cloud

Manufacturing. Robotics and Computer-Integrated

Manufacturing,75-86.

http://dx.doi.org/10.1016/j.rcim.2011.07.002

[2] Svantesson, D. and Clarke, R. (2010) Privacy and

Consumer Risks in Cloud Computing. Computer Law &

Security M. Maqableh et al. 199 Review, 391-397.

http://dx.doi.org/10.1016/j.clsr.2010.05.005

[3] Jadeja, Y. and Modi, K. (2012) Cloud Computing—

Concepts, Architecture and Challenges. 2012 International

Conference on Computing, Electronics and Electrical

Technologies (ICCEET).

http://dx.doi.org/10.1109/ICCEET.2012.6203873

[4] Malathi, M. (2011) Cloud Computing Concepts. 3rd

International Conference on Electronics Computer

Technology (ICECT).

[5] Nexogy (2014) The Impact of Cloud Computing for

VoIP. http://nexogy.wordpress.com/2011/09/27/the-impact-

of-cloud-computing-for-voip/

[6] Arshad, J., Townend, P. and Xu, J. (2013) A Novel

Intrusion Severity Analysis Approach for Clouds. Future

Generation Computer Systems,416-428.

http://dx.doi.org/10.1016/j.future.2011.08.009

[7] Modi, C., et al. (2013) A Survey of Intrusion Detection

Techniques in Cloud. Journal of Network and Computer

Applications,42-57.

http://dx.doi.org/10.1016/j.jnca.2012.05.003

[8] Subashini, S. and Kavitha, V. (2011) A Survey on

Security Issues in Service Delivery Models of Cloud

Computing. Journal of Network and Computer

Applications,1-11.

http://dx.doi.org/10.1016/j.jnca.2010.07.006

[9] Rabai, L.B.A., et al. (2013) A Cybersecurity Model in

Cloud Computing Environments. Journal of King Saud

University—Computer and Information Sciences, 63-75.

http://dx.doi.org/10.1016/j.jksuci.2012.06.002

[10] Karajeh, H., Maqableh, M. and Masa’deh, R. (2014)

Security of Cloud Computing Environment. 23rd IBIMA

Conference on Vision 2020: Sustainable Growth, Economic

Development, and Global Competitiveness.

[11] Mathur, P. and Nishchal, N. (2010) Cloud Computing:

New Challenge to the Entire Computer Industry. 2010 1st

International Conference on Parallel Distributed and Grid

Computing (PDGC), Solan, 28-30 October 2010, 223-228.

[12] Maqableh, M., Samsudin, A. and Alia, M. (2008) New

Hash Function Based on Chaos Theory (CHA-1). 20-27.

[13] Maqableh, M.M. (2010) Secure Hash Functions Based

on Chaotic Maps for E-Commerce Applications.

International Journal of Information Technology and

Management Information System (IJITMIS), 12-22.

[14] Maqableh, M.M. (2010) Fast Hash Function Based on

BCCM Encryption Algorithm for E-Commerce

(HFBCCM). 5th International Conference on E-Commerce

in Developing Countries: With Focus on Export, Le Havre,

15-16 September2010, 55-64.

[15] Maqableh, M.M. (2011) Fast Parallel Keyed Hash

Functions Based on Chaotic Maps (PKHC). Western

European Workshop on Research in Cryptology, Lecture

Notes in Computer Science, Weimar, 20-22 July 2011, 33-

40.

[16] Maqableh, M.M. (2012) Analysis and Design Security

Primitives Based on Chaotic Systems for E-Commerce.

Durham University, Durham.

[17] Wikipedia (2014) Cloud Computing. Wikipedia

Contributors.

http://en.wikipedia.org/w/index.php?title=Cloud_computin

g&oldid=616939446

[18] Khorshed, M.T., Ali, A.B.M.S. and Wasimi, S.A.

(2012) A Survey on Gaps, Threat Remediation Challenges

and Some Thoughts for Proactive Attack Detection in

Cloud Computing. Future Generation Computer Systems,

833-851

http://dx.doi.org/10.1016/j.future.2012.01.006

[19] Benny Karunakar, D. and Datta, G.L. (2007)

Controlling Green Sand Mould Properties Using Artificial

Neural Networks and Genetic Algorithms—A Comparison.

Applied Clay Science, 58-66.

http://dx.doi.org/10.1016/j.clay.2006.11.005

[20] Abdella, M. and Marwala, T. (2005) The Use of

Genetic Algorithms and Neural Networks to Approximate

Missing Data in Database. Computing and Informatics,

577-589.

[21] Fraile-Ardanuy, J. and Zufiria, P.J. (2007) Design and

Comparison of Adaptive Power System Stabilizers Based

on Neural Fuzzy Networks and Genetic Algorithms.

Neurocomputing, 2902-2912.

http://dx.doi.org/10.1016/j.neucom.2006.06.014

[22] Kumar, P. and Verma, A. (2012) Scheduling Using

Improved Genetic Algorithm in Cloud Computing for

Independent Tasks. ICACCT12, Chennai, 3 November

2012.

[23] Goñi, S.M., Oddone, S., Segura, J.A., Mascheroni,

R.H. and Salvadori, V.O. (2008) Prediction of Foods

Freezing and Thawing Times: Artificial Neural Networks

and Genetic Algorithm Approach. Journal of Food

Engineering,164-178.

http://dx.doi.org/10.1016/j.jfoodeng.2007.05.006

[24] Group, L. (2014) Optimisation of Collector Form and

Response.

http://www.engineering.lancs.ac.uk/lureg/group_research/w

ave_energy_research/Collector_Shape_Design.php.

[25] Heckerling, P.S., Gerber, B.S., Tape, T.G. and Wigton,

R.S. (2004) Use of Genetic Algorithms for Neural

Networks to Predict Community-Acquired Pneumonia.

Artificial Intelligence in Medicine, 71-84.

[26] Varahrami, V. (2010) Application of Genetic

Algorithm to Neural Network Forecasting of Short-Term

Water Demand. M. Maqableh et al. 200 International

Conference on Applied Economics—ICOAE, Athens, 26-28

August 2010, 783-787.

[27] Chen, C.R. and Ramaswamy, H.S. (2002) Modeling

and Optimization of Variable Retort Temperature (VRT)

Thermal Processing Using Coupled Neural Networks and

Genetic Algorithms. Journal of Food Engineering, 209-

220.http://dx.doi.org/10.1016/S0260-8774(01)00159-5

[28] Tadiou, K.M. (2014) The Future of Human Evolution.

http://futurehumanevolution.com/artificial-intelligence-

future-human-evolution/artificial-neural-networks

[29] Li, L.Q. (2009) An Optimistic Differentiated Service

Job Scheduling System for Cloud Computing Service Users

and Provider. 3rd International Conference on Multimedia

and Ubiquitous Engineering, Qingdao, 4-6 June 2009, 295-

299.

[30] do Lago, D.G., Madeira, E.R.M. and Bittencourt, L.F.

(2011) Power-Aware Virtual Machine Scheduling on

Clouds

Using Active Cooling Control and DVFS. 9th International

Workshop on Middleware for Grids, Clouds and e-Science,

Lisboa, 12-16 December 2011.

[31] Dutta, D. and Joshi, R.C. (2011) A Genetic-Algorithm

Approach to Cost-Based Multi-QoS Job Scheduling in

Cloud Computing Environment. International Conference

and Workshop on Emerging Trends in Technology (ICWET

2011)- TCET, Mumbai, 25-26 February 2011.

[32] Ghanbari, S. and Othman, M. (2012) A Priority Based

Job Scheduling Algorithm in Cloud Computing. Procedia

Engineering, 778-785.

[33] Xu, B.M., Zhao, C.Y., Hu, E.Z. and Hu, B. (2011) Job

Scheduling Algorithm Based on Berger Model in Cloud

Environment. Advances in Engineering Software, 419-425.

http://dx.doi.org/10.1016/j.advengsoft.2011.03.007

[34] Li, K., et al. (2011) Cloud Task Scheduling Based on

Load Balancing Ant Colony Optimization. 6th Annual

China Grid Conference, Dalian, 22-23 August 2011.

[35] Morariu, O., Morariu, C. and Borangiu, T. (2012) A

Genetic Algorithm for Workload Scheduling in Cloud

Based E-Learning. Proceedings of the 2nd International

Workshop on Cloud Computing Platforms (CloudCP 12),

Bern, 10 April 2012.

The Future Roadmap of Artificially Intelligent Electric Vehicles

Sanjay Devkishan Taneja

Assistant Professor

Amity Institute of Information Technology

Amity University Lucknow Campus

Lucknow, India

sdtaneja@lko.amity.edu

Abstract— This paper presents an ongoing overview of much

awaited emergence of electric vehicles (EV), its background, market

trends and acceptance for these electrically run vehicles in different

countries including India, technological developments towards

Artificial Intelligence powered Electric Vehicle, perception of

government and our society for these vehicles followed by major

challenges. The limited natural stock of Petrol, Diesel,the domination

of countries naturally blessed with enormous fuel repository over

the rest of the world, the observed climatic changes due to global

warming and serious pollution issues are the few valid reasons which

has encouraged the inception and development of electrical vehicles

worldwide.

The highest source of emission of Green House Gas from human

based activities in major countries of the world is from surface

transportation and other related means as per the data collected in

’2010’. The emissions from this sector are expected to escalate by

approximately 33% from 1990 to 2030 on account of predicted

growth rate in the number of vehicles on Road.

Aim of the study is to look upon the latest and upcoming

developments in the market for the electric and AI-electric vehicles, ,

their mutual comparisons and finally to know the outlook and the

support of Government for this new emerging sector.

Keywords—Electric Vehicle, EV Market, AI-Electric Vehicle,

Trends and Challenges

I. INTRODUCTION

The limited natural stock of Petrol, Diesel and CNG, the

supremacy of countries naturally blessed with enormous fuel

repository over the rest of the world, the observed changes in

the direction of global warming and serious pollution issues

are the few valid reasons which has encouraged the inception

and development of electrical vehicles worldwide.

To join hands with the ambitious new 2030s Carbon

reduction targets defined by UK and to account for

increasing level of customer awareness for various

environmental related issues, a large scale of R&D has been

done and different propulsion methods are discovered and

developed.

But carbon dioxide is not the only greenhouse gas: methane,

nitrous oxide and numerous others also trap heat as they

accumulate in the atmosphere, and therefore must be included

the list. The highest source of emission of Green House Gas

from human based activities in the major countries of the

world is from surface transportation and other related means

as per the data collected in ’2010’ [2]. The emissions from

this sector are expected to escalate by approximately 33%

from 1990 to 2030 on account of predicted growth rate in

the number of vehicles on Road [1].

The purpose of this report is to describe the roadmap of

electric vehicles, latest scenario & technological developments in AI powered electric vehicles, perception of government and our society for these vehicles followed by major challenges.

II. HISTORY OF ELECTRIC VEHICLE

An electric vehicle works with a single or dual electric motors

for propulsion. It may be powered a collector system by

electricity from off-vehicle sources. EVs mainly covers

surface and underwater vessels.

The electric vehicle cannot be regarded as the recent

development. It has been around for over 100 years and the

journey with technological changes goes on to the present.

England and France were amongst the few who actually

developed the electric vehicle in the late 1800s.

The attention and contribution for such electric vehicles came

from Americans in the year 1885. Many innovations were

followed but the actual interest was seen increasing greatly in

the late 1890s and early 1900s[4].

Sudden plunge in the cost of the batteries in the Chinese

market, growing commitments from major auto makers,

strong policy support from the governments and considerably

very low operational costs have placed electric vehicles

(EVs) on the track to smoothly override gasoline-powered

vehicles. Incredibly, the United States electric vehicle have

registered a growth on an average of over 32% annually from

the year 2012 to year 2016 and 46% over the year ending in

June 2017.

Recently in February 2018, Twenty Two Motors launched it

smart electric scooter, Flow, equipped with AI and cloud

capabilities at the Auto Expo 2018 being held at Greater

Noida and New Delhi.

The Government of India has been pushing for a shift towards

electric vehicles, and Flow along with other electric vehicles

launched at the expo seem to be a step towards this shift.

III. LATEST TRENDS IN ELECTRIC AND AI - POWERED

ELCTRIC VEHICLES

There are a variety of alternative fuel based modes of surface

transportation. However, amongst all those, electric vehicles

have been on the government radar nationally and

internationally.

As per the year 2015 Global Automotive Executive Survey

conducted by KPMG International , less than one in 20

vehicles is expected to be equipped with electrified

powertrains by the year 2020. Only 0.01 out of 100 cars are

expected to be equipped with fuel cells which comes out to be

about 16,000 units per annum by 2020 [4].

The growth rate of the number of electric vehicles traded

each year globally is mounting swiftly from 45,000 units in

the year 2011 to 300,000 plus units in year 2014 as per the

data in Figure-1, showing Annual global electric vehicle

sales..

Figure-1 : Annual global electric vehicle sales.

The worldwide electric vehicle cumulative registrations

matured notably at a 92% CAGR to touch the expected count

of 665,000 electric cars on road, by the fall of year 2014.

Although US has the largest share with the world’s biggest

fleet of e-vehicles, China has emerged as one of the best

upcoming market for all types of electric vehicles with 36,500

electric buses , incredibly 230 million electric bikes, 83,000

electric cars to be on road by the year 2014.

Honda's latest concepts focus on artificial intelligence only.

Honda has unveiled a series of AI devices and transport

concepts at this year's Tokyo Motor Show, including a sports

car that communicates with its driver.

Figure-2 : National Electric Vehicle Sales Goals for 2020-30

The British and French governments have recently said they

will ban sales of gasoline- and diesel-powered vehicles by

2040.

China is also tightening regulations to encourage the adoption

of electric cars as per the data in Figure-3, showing The Rise

of Electric Cars.

Figure-3 : The Rise of electric cars.

All these conclude that a new world order is coming, driven

by AI and electric cars.

A. Highlights of AI powered electric vehicles

Well known Twenty Two Motors recently launched its

smart e-scooter with AI and cloud capabilities at the Auto Expo 2018 organized at Greater Noida and New Delhi.

The scooter will have the capability of experiential learning , learning of the personal driving etiquettes and behavior of all its regular drivers , enabling them to customize the driving attributes such as , the pickup speed, torque as per heir personal needs.

Audi's AI-CON is designed to run over 700 kilometers on a single charge of just 30-40 minutes. The rapid advancement in the field of self-driven cars guarantees very low fatal human errors on the road and will also ease severe traffic congestions.

Very recently, an AI-powered electric vehicle has been shaped with a bundle of technologies viz. digital image processing for the purpose of road detection, obstacle detection and Lidar, Sonar and Infrared based Obstacle avoidance.

Toyota Research Institute is planning to budget $36 million over the next four years toward deploying artificial intelligence to work on identifying suitable materials that can be readily approved for batteries or catalysts that power hydrogen-fueled cars.

Mitsubishi Electric has confirmed that its AI-powered camera systems will obsolete car mirrors, and much more is happing in this direction all over the world.

B. Indian Government perception for electric vehicles with

Artificial intelligence

The Indian Government has been pushing for a big shift

towards AI enabled electric vehicles. With the Indian Government’s inclination to declare India a 100% EV nation by the year 2030, as indicated in Figure-4.

Figure-4 : Scenario of Electric Vehicles in India

This push for EVs will create a remarkable change in Country’s energy security priorities, securing lithium supplies,

a key raw material for making low priced and high life batteries.

C. Other country initiatives in EV space

With Norway in the worldwide lead, Europe has been set for the race in electric vehicles production. Norway as on today has the highest per capita number of electric vehicles in the globe : nearly 100,000 in a country of 5.2 million population.

The ambitious emissions-reduction targets has aimed growing support from government and the various auto led industries for the rapid technological advancement in electric vehicle zones all over the Europe.

China has been ranked first in the world in electric vehicles production and usage with over 600,000 electric vehicles on the roads , targeting over 5 million by the year 2020. The United States globally ranks third in the row with almost 500,000 electric vehicles registered as on date.

D. Major challenges in electric vehicles

The only major challenge is the vehicle cost as electric vehicle battery is expensive. The batteries in electric cars are supposed to be able to hold huge amounts of charge to make the cars practical on road. They have to be designed and built using high quality expensive raw materials, most of which are tough to procure. The charging stations are another the big challenge. Recharging time is high and roughly consumes around 7-8 hours which seems to be unfeasible during business hours.

IV. CONCLUSION

As per the need of the time, Electric Vehicles are also

proven one of the safest and best option for surface

transportation. By this report, futuristic trends and AI

application has to increase in Indian market as well for the

acceptance of electric vehicle by a common man. By this

we also came to know about current key technological

development in the field of electric vehicles in global

market.

Presently, China and most of the European countries and

then US has registered maximum percentage sale of

electric vehicles as compared to other countries of the

globe. We also studied the inclination of government as

one of the major support for increasing sale of electric

vehicles in various countries.

We also learned the valid reasons why Indian EV market is

still far behind in this increasing race of electric vehicles

specially in the launch of vehicles with AI. What is our

forevision by the year 2030 has also been studied as

regards to design and technological developments in the

emerging area of electic vehicles.

REFERENCES

[1] Neilsen Ole Kenneth and Neilsen Malene, “Projections of Green House

Gas Emission 2010 to 2030” National Environmental Research Institute, Aahrus University, vol. 1, pp. 9-10, September 2011.

[2] GreenHouse Gas Emission (2018) from https://www.epa.gov/ghgemissions/sources-greenhouse-gas-emissions.

[3] Mader, Jerry, and Transportation Energy Center TEC. "Battery powered vehicles : Don’t rule them out." (2006).

[4] Edison Tech Centre (2018) from http://www.edisontechcenter.org/ElectricCars.html

[5] Study by KPMG ‘Emerging trends and technologies in the automotive sector- Supply chain challenges and opportunities’ K- Klynveld, PPeat,

[6] Trends in vehicle concept for Journal of International Battery, Hybride fuel cell Electric vehicle symposium on nov-2013, issu 1, volume

[7] Hybrid and Fuel Cell Electric Vehicle Symposium & Exhibition (EVS-25), Shenzhen, China. 2010.

[8] Liu, Jinsong. "Research on the Development Strategies of New Energy Automotive Industry Based on Car Charging Stations and Battery

[9] Weinert, Jonathan, et al. "The future of electric two-wheelers and electric vehicles in China." Energy Policy 36.7 (2008): 2544-2555.

[10] Wirges, Johannes, Susanne Linder, and Alois Kessler. "Modelling the development of a regional charging infrastructure for electric.

MATLAB Based: Automatic Voice Assistance System for Visually Impaired

Anupam Bhardwaj

Department of CS&E

Amity School of Engineering

and Technology

AUUP,Lucknow

bhardwajanupam1@gmail.c

om

Ankit Yadav

Department of CS&E

Amity School of Engineering

and Technology,

AUUP,Lucknow

yadavank80@gmail.com

Dr. Geetika Srivastava

Department of E&CE

Amity School of Engineering

and Technology,

AUUP,Lucknow

gsrivastava2@lko.amity.edu

Introduction From a long period of time reading and

detecting the sign boards, labels and other

surrounding objects is a matter of great

concern for visually impaired people. Since

the sign boards and posters can’t be written in

braille script which is easily readable and

understandable by the sightless people, so

there is a need to solve this problem. This

project focuses on this problem and is

designed to solve this problem without any

issues. The text written on the sign board or

posters is detected by the program and it is

provided as a voice assistance to the person

through an ear piece.

Working The image of the sign board or poster is

captured by the camera that is mounted on the

spectacles of the blind person. This image is

then transmitted through Bluetooth to the

persons mobile phone. This image is then

processed to detect the MSER (Maximally

Stable Extremal Region) region. MSER

region is that region which have very less

variation in intensity. These regions are

detected by performing stability test using

Intensity Threshold Level. The detected

MSER region is then processed for their stroke

width transformation. A stroke in an image is a

region of continuous band having nearly

constant width. Starting by assuming infinite

width, we start to increment through the region

of one edge to another in the direction of

calculated gradient. The direction of gradient

is mostly directed toward the edge. Finally, the

pixels that have similar stroke width and

satisfying certain conditions are grouped

together.

After stroke width transformation the image is

then send for filtration. In the filtering process

certain filtering condition are taken into

account. For an image to satisfy filtering

condition aspect ratio must be in small range.

This is done in order to filter long and narrow

components. Components with too large or

small are ignored. Finally, we calculate the

boundary box is calculated for each character.

These boundary boxes are then connected to

form a complete word or a sentence. These

detected words and sentences are the

converted from text to speech and the audio is

played through the speakers

Flow Diagram

Future Scope Currently this project is only able to detect text

form an image taken using a webcam or an IP

camera. But in future it will be able to detect

text as well surrounding objects using a

Bluetooth camera mounted on the spectacle of

the sightless person and the processing will be

done through an application on the persons

smartphone. Apart from only detection this

project will also be able to give translation of

the detected text in various languages.

Conclusion This project will help a lot of visually impaired

people around the world by making them more

independent. It will allow them to work freely

and live in this world fearlessly. This project

demonstrated in not only feasible for

technology but has a large market potential. As

our aim is to make this world a better place to

live for everyone.

References:

1. Ezaki, Nobuo, Marius Bulacu, and

Lambert Schomaker. "Text detection

from natural scene images: towards a

system for visually impaired

persons." Pattern Recognition, 2004.

ICPR 2004. Proceedings of the 17th

International Conference on. Vol. 2.

IEEE, 2004.

2. Do, Cuong, and Michael Merman.

"Automatic bank teller machine for

the blind and visually impaired." U.S.

Patent No. 6,061,666. 9 May 2000.

3. Cohen, Albert E., et al. "Method and

system for assisting the visually

impaired in performing financial

transactions." U.S. Patent No.

6,464,135. 15 Oct. 2002.

4. Moore, Michael C., et al. "Device for

assisting the visually impaired in

product recognition and related

methods." U.S. Patent No. 5,917,174.

29 Jun. 1999.

5. Garaj, Vanja, et al. "A system for

remote sighted guidance of visually

impaired pedestrians." British Journal

of Visual Impairment 21.2 (2003): 55-

63.

6. Sanders, Aaron D., and Kasey C.

Hopper. "System for indoor guidance

with mobility assistance." U.S. Patent

No. 9,539,164. 10 Jan. 2017.

7. Zientara, Peter A., et al. "Third Eye: a

shopping assistant for the visually

impaired." Computer 50.2 (2017): 16-

24.

8. Gowtham, S. U., et al. "Interactive

Voice & IOT Based Route Navigation

System For Visually Impaired People

Using Lifi." (2017).

9. Rahman, Samiur, Sana Ullah, and

Sehat Ullah. "Obstacle Detection in

Indoor Environment for Visually

Impaired Using Mobile

Camera." Journal of Physics:

Conference Series. Vol. 960. No. 1.

IOP Publishing, 2018.

10. Ani, R., et al. "Smart Specs: Voice

assisted text reading system for

visually impaired persons using TTS

method." Innovations in Green Energy

and Healthcare Technologies

(IGEHT), 2017 International

Conference on. IEEE, 2017.

11. Engelke, Robert M., and Kevin R.

Colwell. "System for text assisted

telephony." U.S. Patent Application

No. 15/010,199.

The Wider Realm to Artificial Intelligence in

International Law Abhivardhan

Amity University Uttar Pradesh, Lucknow

Malhaur Near Railway Station, Nizampur, Gomti Nagar Extension,

Lucknow, Thasemau, Uttar Pradesh, India, 226010 abhivardhan2@gmail.com

Abstract— International Law subsists towards a traditionalist

approach of the legal instruments existing between states. This

traces its origins towards a sui generis act of responsive

considerations, which form the psychological and material

content of state practice. However, the basic tenets of

International Law have not yet developed. The limiting factors

that have prevented it are not exhaustive nor exclusive but just

heavily traditional. A better set of examples may include the

jurisprudence of Budapest Convention on Cybercrime and the

action against Russia by NATO in 2007 regarding the cyber-

attack in Estonia. Moreover, the approaches of North Korea

pertaining to WannaCry and the need of a civilized set of legal

principles are in scarcity or if there are any such, the

traditionalism of the so-called manifestation is in a permanent

question of concern, which, in the end, is fatal for the future

approaches in International Law The question of individual

agility and representation in a cyber realm stands unquestioned

by International Law even where there is a subjected realization

of AI, based on null human intervention, which may broaden.

Artificial Intelligence does not only manifest the sui generis realm

of nullity of human sense and presence, but it also moves towards

the synthesis towards a more technical but reflective aspect of

society, which shall affect statehood and non-state actors. The

paper thus focuses on the ad hoc purpose to devise some basic

principles on how the instruments of International Law can

modernize its limits via AI.

Keywords— International Law, Artificial Intelligence, Political

Subjectivity, Informational Bias, Sovereignty.

I. INTRODUCTION

International Law is considered to be the concept of legal

reflection among ‘civilized nations’, which is true in its own

origins. However, it traces back its very nearest vicinities in

the Roman Empire, where state sovereignty and statehood

had only a mere relevance up to absolutism, obedience and

deterrence. There, in the contemporary era, it widens and

we seek the formation of International Humanitarian Law,

the 1899 and 1907 Hague Conventions and the formation of

United Nations (also the League of Nations, which

however, had least value of concern). Well, increasing

global integration is not an entirely new phenomenon,

contemporary globalization has had an unprecedented

impact on global public health and is creating adverse

challenges for international law [1] and policy including the

municipal laws and legal system of various developed and

developing nations. However, machines matching humans

in general intelligence (possessing common sense and an

effective ability to learn, reason, and plan to meet complex

information-processing challenges transversely a wide

range of natural and abstract domains [2]) have been

expected since the invention of computers in the 1940s. [2]

Yet starting at a sub-human level of development and

enduring throughout all its succeeding development into a

galactic superintelligence observable, the AI’s conduct is to

be guided by an essentially unchanging final value that

becomes better understood by the AI in direct significance

of its general intellectual progress- and likely quite

differently by the mature AI than it was by its original

programmers, though not different in a random or hostile

way [2] but in a compassionately appropriate way.

Henceforth, wherever AI substantiates its subjected matter

of relevance, it enshrines a sui generis relationship with the

limits and phases of human intervention that it seeks.

However, when it comes to International Law, the concept

itself is too stringent. It is far traditionalized, which does not

necessitate those basic approaches that we require still.

Since, the conglomeration of AI and International Law

at the level of nations is not a reality yet, therefore there

does not exist any real case. However, in the presence of

some hypothetical considerations, it is asserted that the

purpose of this paper to give newer directions to guide

towards a more better and understandable theoretical and

applied reality that we can achieve via AI to bring the

applied mechanisms of International Law towards

modernity. A good example is the concept of sovereign

recognition, which is dealt differently in the paper. More

examples can be regarding Artificial Diplomacy, where

human intervention to the political aspect of diplomacy is

attained but there exists a watcher to control political

subjectivity as only a mere human-oriented data and not a

directive. This is widely discussed further in the paper.

Therefore, this paper gives an earnest effort and

approach to conglomerate the basic principles and

conclusions that can guide the jurisprudence of

International Law towards the next level of human and non-

human configurations irrespective of the proclamation that

the assertions may not become an absolute truth.

II. THE TRADITIONAL SUBJECTIVITY OF INTERNATIONAL LAW

The basic foundation of International from the Roman

Law favor sovereignty more and remain either moot or

restrictive over the relevance of self-determination, which

is the baseline to understand what International Law is.

Even though the introduction of International Law is

benefitted by the erga omnes of every nation, relations in

the international level ascribe a different setup of how these

principles ascribe the general relevance of the sovereigns

and their functionary manifestations and establishments.

Thus, this introduction shall focus a little on the divine

theory of state under the Roman Law, which was advocated

at its best by Alberico Gentili.

Gentili stipulated, “Sovereignty is absolute and

perpetual power… This sovereignty means that the prince

never finds anything above him, neither human being, nor

law… This power is absolute and without limitation… That

the ‘Prince is not bound by law’ is law, as is also that ‘Law

is what pleases the prince’. And this is no barbarian law but

Roman law, the first and foremost among human laws…

And so, what is called regal prerogative in England…is

absolute power [3]”.

However, the virtue of notions does not have any

substantiated relevance and existence when we deal with

the contemporary principles. One of them, excellently

states- “It is a principle of international law, and even a

general conception of law, that any breach of an

engagement involves an obligation to make reparation [4]”.

Still, this is just based on the law of obligations, wherever

the principle that a State cannot rely on its own wrongful

conduct to avoid the consequences of its international

obligations is capable of novel applications and

circumstances can be imagined where the international

community would be entitled to treat a new State as existing

on a given territory, notwithstanding the facts [5]. Thus, if

we consider the mere concept of consent as per the Vienna

Convention on Law of Treaties [6], we come to know that

the convention merely subjects better leverage to

sovereigns, which if we understand in more general aspects,

we can understand that still, obligatory limitations in the

creativity to embody the realism from the International

Relations theory is sculpted, which is reflected in the

International Covenants of 1966 and the Geneva

Conventions of 12 August 1949. Irrespective of that,

International Law converts into a traditional system of

action, which is due to the traditional attributes of the state’s

understandable in the state practices and the customary

international law that makes it jus cogens. Still, the attacks

by Russian hackers to access the information of the website

owned by the Government of Estonia is joined as a violation

of the Article 2 of the Charter of the UN and so the United

States invoked the Article 51 of the Charter and the Article

5 of the North Atlantic Treaty to intrude into a cyber-war by

Russia. Moreover, the actionable constraints created by

Stuxnet and the required fear understandable from North

Korea’s WannaCry leaves the International Community in

a disarray to substantiate how the cyber realm is lead. Well,

this is not the case of cybercrimes at an international level

as deterrence and retribution is just the means to limit and

regulate the cyber community when it is subjected to least

control in the vicinity of its young age. Hence, this must be

concluded-

• It is not the fault or blame of the cyber realm to

exist, but that of the understanding and action that

law has primarily dealt;

• Responsiveness and deterrence is a need, but that

does not recognize the realism and informational

relevance that the cyber realm is manifesting;

• Traditionalizing legal instruments is a drawback to

clearly regulate a societal realm, whether is it

human or non-human

III. THE REASON WHY AI CAN GUIDE INTERNATIONAL LAW

Artificial Intelligence in all its sui generis aspects, is the

realm of subjected limitations to human instruments’

intervention and represents a general realm of self-developed

attributes of recognition, representation and participation,

which, in the end, necessitates the broader outlook of the lives

that non-human and human instruments make up with the

coherence of furtherance.

Well, the nexus can never be easy as both the disciplines are

extremely divergent and could never have observed the basic

relationship, which is to be expected. Now, Vienna Convention

on Law of Treaties [6]. However, even after the successful

construction of an operational rule-base, the meaning of the

individual rules remained a mystery [7]. In International Law,

we generally develop entities, within the ambit of ‘what is there

actually’ and ‘how that can be relevantly persisted’. Hence,

where operational configurations build up the required set of

contingencies, we come up with more clearer solutions.

However, in International Law, the subjectivity problem can

be demarcated if we change the prioritization of International

Relations with regards to the factor of utility. It is actually

possible, which is greatly conceivable.

Let us take an example. The Russian Federation in 2007,

attacked the cyber systems of the State of Estonia, a European

nation. A series of cyber-attacks began 27 April 2007 that

swamped websites of Estonian organizations, including

Estonian parliament, banks, ministries, newspapers and

broadcasters, amid the country's disagreement with Russia

about the relocation of the Bronze Soldier of Tallinn, an

elaborate Soviet-era grave marker, as well as war graves in

Tallinn [8] . It was established to what has become known as

“Web War 1”, a concerted denial-of-service attack on Estonian

government, media and bank web servers that was precipitated

by the decision to move a Soviet-era war memorial in central

Tallinn in 2007 [9]. Similarly, on 20 July 2008, weeks before

the Russian invasion of Georgia, the "zombie" computers were

already on the attack against Georgia [10] [11]. The website of

the Georgian president Mikheil Saakashvili was targeted,

resulting in overloading the site. The traffic directed at the Web

site included the phrase "win+love+in+Rusia". The site then

was taken down for 24 hours [12] [13] . Nothing compares to

the destructive power of a nuclear blast. But cyber-attacks

loom on the horizon as a threat that is best understood as an

extraordinary means to a wide variety of political and military

ends, many of which can have serious national security

ramifications. [14]

Now, sovereignty and self-determination cannot be non-

human or non-human intervened. Why? Sovereignty as a

political concept is only for humans. That is why our leaders

can maintain the concept’s fragility and regulate their

representation. However, what if became possible that

sovereignty is more federalised? It means that each and every

aspect of it is under a practical recognition. AI can possibly do

that. However, the AI itself cannot dominate human interests

and opinion, which is the matter of necessitated caution.

Obviously, all the political realms are human and till human

thought and presence is a reality, sovereignty as a matter of

relevance must be as it is. So, by this example, we can

understand that International Law, may become a little

intervening but that would realise into a very dangerous aspect

sometimes. While machine learning is commonly associated

with automated driving, search algorithms, and image

recognition, it extends to cover the law. [15] Still, this may not

be that dangerous for any matter of concern as such, which is

related to some very basic concerns. If we go towards the

ambiguity of customary international law, we can solve its

problem by maintaining a recognizable limit of human

intervention and present relevance, which is itself maintained

by the AI configurations.

IV. THE RELATIVITY FROM AI TO THE INSTRUMENTS OF

INTERNATIONAL LAW AND RELATIONS

Artificial Intelligence can provide a wider utility to the

applied realms of International Law, that is, global

cooperation in IEL and IHL, recognition of cyber realm and

its users with a wider realization of the relevance of every

action in the purview of it and others, which are still

unknown to us. The notion of control, it can be assumed,

has garnered support, because it is a vague term. It is a

longstanding practice in international law to use vague legal

terms which all parties can construe according to their

preferences when positions diverge widely. Thus, the

U.S.A. and other states likely to have autonomous weapons

systems may interpret control loosely. This allows them to

see as lawful a wide variety of constellations, such as an

operator who simultaneously controls hundreds of

weaponized swarm drones or an armed underwater or air

vehicle which has been broadly authorized to use force and

loiters for months [16]. Here are a few examples and their

analysis per se for the cause of the required objective to bear

a clarity.

A. Global Health Law

Global Health Law ascribes more than a mere reflective

action under globalization and state responsibilities and

intervention of International Organizations. Thus, referring

Lawrence Gostin, “Global health law is a field that

encompasses the legal norms, processes, and institutions

needed to create the conditions for people throughout the

world to attain the highest possible level of physical and

mental health. [17]” It is widely recognized that current

system of global health governance is insufficient to meet

the wide range of challenges and opportunities brought by

globalization [18]. A fundamental challenge of global

health governance is the state-centric nature of international

law [17], where a critical limitation of the state-centric

nature of international law is its inability to incorporate

nonstate actors in the legal framework for global health

governance. [17] In contemporary global health

governance, states are apparently unwilling to develop

international legal instruments that create binding and

meaningful obligations and incentives and provide deep

funding or services for the protection of the world’s poorest

people. As a consequence of the voluntary nature of

international law and the overriding principle of

sovereignty, states have established only a limited legal

framework for national action and international cooperation

to advance domestic and global public health [17], which

defies the principles of IHRL. Thus, just maintaining a

vague relevance of the concept of sovereignty would not

benefit the realm. Instead, via the instrumenting

demarcation of AI, we can actually, make better, helpful and

cogent responsible drives among states. A great so-called

resemblance is the Earthwatch, initiated by the UNEP in the

mandate of the General Assembly, where governments are

put on one table to negotiate and effectively avail

themselves of watching the contingencies arisen per se.

Here, AI is not used. Nevertheless, even if AI comes into

play, that would never defy state sovereignty because states

give consensual protection, where treaty obligations are

actually protected by the virtue of AI. Second, the AI shall

provide a close watch to the anomalies to solve all the

required problems via keeping in mind the sovereign

obligations and legal backing and no so-called “politically

unusual” interests. Third, as International Law cannot be

automated, interests shall be more transparent because even

negotiations arise into something very secretive or abstract,

that would be made more concretely analyzable. The

creation of international legal norms, processes and

institutions provides an ongoing and structured forum for

states to develop a shared humanitarian instinct on global

health. But the problem of using international law as a tool

for effective global health governance has long perplexed

scholars, and for good reason. [17] Thus, let us give an

extended but extensively ascertainable utility to states for

not opening up their interests but their activities among

states, when they are required.

B. Customary International Law

Customary International Law is the vaguest idea in

International Law. It is criticized for the demarcated

categories that it has- that is, the psychological element and

the material element. Thus, if we understand how it goes on,

let us analyze one important consideration by the ICJ in

Hungary v. Slovakia [19]

The Court analyzed one statement submitted by Slovakia

pertaining to a Variant C, which was- “It is a general

principle of international law that a party injured by the non-

performance of another contract party must seek to mitigate

the damage he has sustained. [19]” The statement, however,

failingly signifies that “while this principle might thus

provide a basis for the calculation of damages, it could not,

on the other hand, justify an otherwise [19]”. Hence, the

responsibility cannot be adjudged. Here, this is important to

understand that states commit one single mistake - they hide

subjective interests, which is justifiable, but they also hide

transparent reasonability, which is nonetheless, so vague.

State practice, is not only restricted by VCLT or jus cogens,

but also the non-generalization of political subjectivity as

data and not directives. That is why even China adopted a

‘liberal’ perspective in its communist political spectrum.

Even Japan and UK need liberal politics and representation

as they know that sovereignty must never be colourless and

ineffective. Thus, AI can advise, control and solve the

nuances of state practices, where it can become a separate

alternative to the sovereign’s application of mind provided

that the political essence existent in the society does not due

and is left to replenish itself with time.

C. Statehood and Federalism

Sovereignty as a concept has been discussed extensively.

However, if we materialize what sovereignty is, then we can

find it out that there is no sovereignty as a material at all. It is

just evolved as just a manifestation of relevance or

authoritarian subjectivity that law takes as a burden; it does not

give any quantitative solution to any sui generis dispute when

self-determination arises as a basic realm from human

instruments to the sovereign. That is why even a democracy

fails in the USA, India and Europe. But that does not mean we

don’t need democratic institutions. To earn democratic

institutions, we need to make it technically more trustworthy

and clear. No subjective data must be converted into a directive

like perception, emotion, ethnicity, religion and sectarianism,

which Artificial Intelligence, being free from all such massive

data of subjectivity, can ably perform. In fact, it can neutralize

sovereignty and redefine every human’s role in a more better

way to provide subjective independence. Now, even AI can be

dangerous. However, it is the endeavor that should keep going.

Here is a simulating exemplification of how the mechanisms

can be initiated to enhance the political, social and economic

fate of mankind, which shall affect the obligations of

International Law. This example is based on to procure AI

systems for diplomatic representations, political assertions and

state polity action mechanisms in accordance with the reality

of the sovereignty of any X nation.

Political independence is guaranteed by the charter of the

UN in its Article 2(3), which treats political element,

information or any other part or constituent of political realms

as data (say). The principles in law and politics are

interpretable. Let us say there shall be two kinds of

interpreters- human and non-human. The non-human

interpreter shall, under the virtue of AI, be given a sui generis

virtue to equate and equitize the requirement, procurement and

limitation of interpretation. However, interpretation in

International Law is human-oriented, i.e., it consists of

application of human virtue and mind, where the capability of

humans and referential presence as curation makes up the task

of interpretation. Now, for politics, this can be a good argument.

However, we can do one more thing. If an AI interface is

acknowledged with all the data or the information converted

into the specifically capsuled data such as the GDP stats,

economic policies and networks at internal and international

level, defence information, legal instruments such as a

constitution, federal laws, codes, statutes, case laws (with

ample limited and suggestive relevance), etc., then we can

surely achieve the cause per se. Provisional results from

Myanmar’s first census in 30 years released over the weekend

show that the country has nearly 9 million fewer people in it

than originally thought, as rights groups decry the absence of

data recognizing its oppressed Muslim Rohingya population.

According to the provisional results, Myanmar now has a

population of 51.41 million, falling short of the estimated 60

million previously believed to be living in the country that was

once all but closed off to the world. [20] This is not in the

purview of International Law directly but shall affect

International Relations. Thus, it shall always have a less

distinctive effect of the utility of international law. Frost and

Sullivan partnered with SocialCops to use our platform to build

a data-driven scorecard comparing each of Johor's 98 police

stations. Frost & Sullivan had already conducted three primary

surveys: Customer Satisfaction Index (CSI) perception, CSI

transaction, and Employee Satisfaction Index (ESI). [21]

Eventually, the dashboard showed what areas felt safest to

citizens, how reliable and fair citizens perceived each police

station, how satisfied police were at each station, and much

more. [21] In 2015, India set out to make its central

government more data driven and transparent. By setting and

tracking clear, quantitative goals, the government would be

able to measure its progress toward building a better India. Niti

Aayog and QCI partnered with SocialCops to make this

possible by unifying and tracking data for all of India's 89

ministries. [22] This, there can be a good possibility.

V. CONCLUSIONS

Nothing is absolute; and so is the AI. It just depends on how

we effectively make up the best human resource in the earth –

data- whether as being material or immaterial, relevant and

connective with nations and other non-state entities.

International Law is not a puppet of the law of treaties and

subjective directives of nations; it is the most beautiful

manifestation of statutory relevance and faith, which is

committed for our countless generations to succeed throughout.

So, if we really need to utilize the resource of AI and its future

for a better international community, we must regulate and

entrepreneur the subjectivity that subjectifies International

Law and just not become a subjective of this narrative and

directive.”

ACKNOWLEDGMENT

The author presents gratitude to his parents and the

colleagues and faculty of Amity University Uttar Pradesh,

Lucknow for moral encouragement and support.

REFERENCES

[1] Lee, K., Buse, K. and Fustukian, S. (eds) (2002). Health Policy in a

Globalising World. Cambridge: Cambridge University Press.

[2] N. Bostrom, Superintelligence. Oxford: Oxford University Press,

2013, pp. 3-210.

[3] A. Gentili, ‘De potestate regis absoluta disputatio’, in Regales

disputationes (London, 1605), pp. 5-11, cited and commented in Panizza, Alberico Gentili, p. 159..

[4] (1928) PCIJ Series A, No. 17, 29. Interpretation of Peace Treaties with Bulgaria, Hungary and Romania, Second Phase, ICJ Reports, 1950, p.

221, 228; Phosphates in Morocco, Preliminary Objections, (1938)

PCIJ, Series A/B, No. 74, 28..

[5] James Crawford, The Creation of States in International Law, 2nd

edition, (Clarendon Press, Oxford, 2006), pp. 477-8. Also see Preamble para 5 of the Israel-Palestine Liberation Organization:

Interim Agreement on the West Bank & the Gaza Strip, signed 28

September 1995, reprinted in (1997) 36 ILM 551 (‘a transitional period not exceeding five years from the date of signing the

Agreement on the Gaza Strip and the Jericho Area on May 4, 1994’).

[6] United Nations, Treaty Series, vol. 1155, p. 331.

[7] Breuker, J. & Wielinga, B. (1989). Models of expertise in knowledge acquisition. In Guida, G. & Tasso, C. (Eds.), Topics in expert system

design, methodologies and tools (pp. 265 -295). Amsterdam,

Netherlands: North Holland.

[8] Traynor, Ian (17 May 2007). Russia accused of unleashing cyberwar to disable Estonia, The Guardian; (1 July 2010). "War in the fifth

domain. Are the mouse and keyboard the new weapons of conflict?".

The Economist. Retrieved 2 July 2010.

[9] Sutherland, Benjamin (2011). The Economist: Modern Warfare,

Intelligence and Deterrence: The technologies that are transforming them, Profile Books, 1-302.

[10] Markoff, John (12 August 2008). Before the Gunfire, Cyberattacks,

The New York Times. Retrieved from http://www.nytimes.com/2008/08/13/technology/13cyber.html.

[11] Wentworth, Travis (23 August 2008). How Russia May Have Attacked Georgia's Internet, Newsweek. Retrieved from

http://www.newsweek.com/how-russia-may-have-attacked-georgias-

internet-88111.

[12] Danchev, Dancho (22 July 2008). Georgia President's web site under

DDoS attack from Russian hackers, ZDNet. Retrieved from

http://www.zdnet.com/article/georgia-presidents-web-site-under-ddos-attack-from-russian-hackers/.

[13] (21 July 2008). Georgia president's Web site falls under DDOS attack, Computerworld, Retrieved from

https://www.computerworld.com/article/2534930/networking/georgia-

president-s-web-site-falls-under-ddos-attack.html.

[14] Kenneth, Geers, (2010). The challenge of cyber attack deterrence,

Volume 26, Issue 3, Computer Law & Security Review, 298-303.

[15] Machine learning and the law is a rather new field. However, the

conference “Fairness, Accountability, and Transparency in Machine Learning” has been organized on a yearly basis since 2014 (see:

http://www.fatml.org). Various conference series and organizations

have also covered the topic broadly for some time, e. g. CEPE (Computer Ethics: Philosophical Enquiry), ETHICOMP, and IACAP

(International Association for Computing and Philosophy). NIPS

(Neural Information Processing Systems) also offered a workshop on Machine Learning and the Law in 2016, organized by Adrian Weller,

Jatinder Singh, Thomas D. Grant, and Conrad McDonnell: see

www.mlandthelaw.org (with papers).

[16] Thomas Burri, “International Law and Artificial Intelligence”, 27

October 2017, available on SSRN: https://ssrn.com/abstract=3060191

(forthcoming in German Yearbook of International Law 2017/18), pp. 1-21.

[17] Lawrence O. Gostin, Allyn L. Taylor; Global Health Law: A

Definition and Grand Challenges, Public Health Ethics, Volume 1, Issue 1, 1 April 2008, Pages 53–63,

https://doi.org/10.1093/phe/phn005.

[18] Dodgson, R., Lee, K. and Drager, N. (2002). Global Health

Governance: A Conceptual Review. Geneva: World Health

Organization and London School of Hygiene and Tropical Medicine

[19] Gabčĭkovo-Nagymaros Project (Hungary v. Slovakia) 1997 I.C.J. 7 (Apr. 9); Chorzow Factory Case (Germany v. Pol.) 1928 P.C.I.J. 47

(ser. A), N° 17 (Sept. 13); Military and Paramilitary Activities in and

against Nicaragua (Nicaragua v. U.S.A.) 1986 I.C.J 14 (June 27); Nuclear Tests Cases (New Zealand v. France) 1974 I.C.J. 4 (Aug. 8).

[20] Heijmans, Philip (September 02, 2014). Myanmar’s Controversial

Census, The Diplomat. Retrieved from https://thediplomat.com/2014/09/myanmars-controversial-census/

[21] CASE STUDY: Evaluating Malaysian Police Stations with Data, SocialCops. Retrieved from https://socialcops.com/case-

studies/crowdsourcing-citizen-feedback-for-malaysian-police/

[22] CASE STUDY: Making India's National Ministries Data Driven,

SocialCops. Retrieved from https://socialcops.com/case-

studies/building-a-system-for-niti-aayog-to-track-89-ministries/

DESIGNING OF LOW POWER AND HIGH PERFORMANCE SAMPLE AND HOLD

CIRCUIT IN 180nm CMOS TECHNOLOGY USING MENTOR GRAPHICS

Snigdha Tripathi1 , Vivek Kumar

2, Dr. Geetika Srivastava

3

Department of Electronics and communication

Amity University, Lucknow Campus sweetsnigii@gmail.com, vkumar1310996@gmail.com , gsrivastava2@lko.amity.edu

ABSTRACT

Objective of the paper is to design low power, high

performance sample and hold circuit in 180nm CMOS

technology. A Switched Capacitor topology being used

for Sample and Hold (S&H) circuit that allows

integration of both digital and analog functions on a

single silicon chip. 180nm CMOS technology with power

supply of 3.3 V is being used. (A/D) converters consists

three parts: (i) sampling (ii) quantization (iii) coding.

The S&H circuit forms fundamental building block for

analog-to digital (A/D) converters and responsible for

sampling and quantization of analog signals. It also

plays an important role in operating speed of (A/D)

converters. S&H circuit helps in reducing the dynamic

errors in ADC at high frequency operations. The

simulation results are based on Mentor Graphics -ELDO

SPICE Simulator. The results of this paper can be a

useful resource for high speed application designers.

I. INTRODUCTION

Sample and Hold circuit is an analogous device

which helps in sampling the continuous varying

voltage of analogous signal and hold its value at

constant level at specified period of time. These

circuits are fundamental analogous memory devices.

These are typically used in Analog to Digital

converters to eliminate distortion in input signal

which disturbs the process of conversion.

Fig. 1 Sample and Hold Circuit

II. BACKGROUND

Figure 1 shows the circuit diagram of basic Sample

and Hold circuit . This circuit consists of switch s0

which is coupled in series with capacitor Cout. The

voltage across capacitor is defined by input Vin and

switch s0 is closed for sampling rate.

Fig 2.Sample and Hold circuit using NMOS

Fig 3. Sampled output of S&H circuit

Figure 2 describes Schematic of basic NMOS

(S&H) circuit. In this switch s0 in figure 1 is

replaced by NMOS transistor. ELDO simulator is

used here for result analysis. The NMOS used is

L=180nm W=720nm Vt=3.3V. Here V1 is the

input analog signal of frequency 10 Mhz and

NMOS trainsistor is used as switching element .V2

is the other input with sampling rate of 100 Mhz and

time period of 100us , pulse =3.3v rise time =0, fall

time = 0,pulse width=50us. C1 is the capacitor of

1pf capacitance. Figure 3 shows the sampled output

of the (S&H) circuit.

When the input clock at gate switches to high and is

switched on , capacitor C1 is shorted to the input. In

this case NMOS is in sampling mode and works in

tracking mode , for that and

.gate to source voltage is

Final equation shows the resistance of NMOS

transistor. M1 is dependent on input voltage and the

resistance results in distortion of hold voltage of

capacitor during hold mode of capacitor.

NMOS transistor is being used for design of (S&H)

circuit instead of PMOS transistor because

A N-channel devices are faster as compared

to P-channel devices as the electron which

are carrier in N-channel device are two

times mobile than holes in P-channel

devices.

Since the impedance of N-channel

transistors are almost as half as the

impedance of P-channel transistors

operating under the identical condition.

Therefore N-channel integrated circuits can

be implemented in less area of silicon

wafer.

PMOS junction area is larger as compare to

N-channel devices and this gives advantage

to N-channel devices. Speed of operation of

metal oxide semiconductor is reduced by

resistor capacitance time constant , thus

capacitance of diode is directly

proportional to the size of diode , which

enhances the speed of N-channel devices as

they consists smaller capacitance.

III. CIRCUIT DESCRIPTION

We have used unity gain amplifier circuit as

requirement of current mode approach. the

major concerns with regard to unity gain

amplifier in these application are their high

frequency response and chip dimesions.

Unfortunately switched capacitor based unity

gain amplifier are often parasitic sensitive.it

used as first stage in various current feedback

amplifiers such as current operational amplifiers.

Fig 4. S&H circuit with unity gain amplifier

Figure 4 shows the circuit of (S&H) of analog to

digital converter. unity gain amplifier circuit is

added to the previous circuit of NMOS based .

As the unity gain amplifier doesn’t amplify but

it provides the gain of 1. Its advantage is that it

isolates the input current of the circuit from

output current of the circuit .unity gain amplifier

circuits comes in two types : voltage follower

and voltage inverter. Here we have used voltage

follower circuit of unity gain amplifier using

CMOS .

IV. RESULTS AND CONCLUSION

Fig.5 Sampled Output with unity gain amplifier

Result analysis of the circuit is simulation based

in mentor graphics – ELDO tool. As the circuit

is designed for low power dissipation while

working on high frequency input signals which

helps the application designers for more

optimization. Figure shows the result of high

gain sampled output of the high frequency

analog input. Result shows the output sampled

voltage following the high frequency analog

input signal which consumes less power which

helps in more optimized circuit designing of

analog to digital converters.

REFERENCES

[1] Venes, Ardie GW, and Rudy J. van-de-

Plassche. "An 80-MHz, 80-mW, 8-b CMOS

folding A/D converter with distributed track-

and-hold preprocessing." IEEE Journal of Solid-

State Circuits 31.12 (1996): 1846-1853.

[2] Shahramian, Shahriar, Sorin P. Voinigescu,

and Anthony Chan Carusone. "A 30-GS/sec

track and hold amplifier in 0.13-μm CMOS

technology." Custom Integrated Circuits

Conference, 2006. CICC'06. IEEE. IEEE, 2006.

[3] Jakonis, Darius, and Christer Svensson. "A 1

GHz linearized CMOS track-and-hold circuit."

Circuits and Systems, 2002. ISCAS 2002. IEEE

International Symposium on. Vol. 5. IEEE,

2002.

[4] Dai, Liang, and Ramesh Harjani. "CMOS

switched-op-amp-based sample-and-hold

circuit." IEEE Journal of Solid-State Circuits

35.1 (2000): 109-113.

[5] Bult, Klaas, and Govert JGM Geelen. "A

fast-settling CMOS op amp for SC circuits with

90-dB DC gain." IEEE Journal of Solid-State

Circuits 25.6 (1990): 1379-1384.

[6] Liu, Ren-Chieh, Kuo-Liang Deng, and Huei

Wang. "A 0.6-22-GHz broadband CMOS

distributed amplifier." Radio Frequency

Integrated Circuits (RFIC) Symposium, 2003

IEEE. IEEE, 2003.

[7] Bernstein, Kerry, et al. "High-performance

CMOS variability in the 65-nm regime and

beyond." IBM journal of research and

development 50.4.5 (2006): 433-449.

Artificial Intelligence in Legal Research and other dimensions of Law

Vagish Yadav

Research Scholar

Amity Law School,

Lucknow, India.

vagish@protonmail.com

Abstract

Legal Research is a time-taking process. It takes

lawyers a lot of patience and hard work to come up

with the right sources and materials for proper legal

decision making. May it be any court of law,

advocates are relevant in any conundrum presented

before the court to solve the respective conundrum.

The use of tools and basic ideas of Artificial

Intelligence can be used to make the legal research

for the lawyer more efficient and problem-solving.

This, would be a better presentation of the case

before the court effecting the legal decision making

positively without any losses to the jobs, which is

mostly thought of when we talk about Artificial

Intelligence.

Use of search methods and planning methods in AI

is used to create such intelligence which can

automatically classify and organize legal text and

summarize the required research at the terms of the

lawyer. This can be done by giving the intelligence

access to all data that shall be used in legal research

for a particular case.

ROSS Intelligence is an example of Artificial

Intelligent Lawyer which does the legal research

efficiently. This is the first step to the subject of

Artificial Intelligence and Law.

I. INTRODUCTION

Artificial Intelligence and Law is a subfield of

Artificial Intelligence(AI). Both the subjects

complement each other and can help each other

develop more. While Artificial Intelligence is in it’s

initial stages to develop into a more enhanced form,

law has already developed it’s base. Legal Problem

solving has been most exposed to the general public

and hence, it is the most worked on. Artificial

Intelligence can take the techniques of problem

solving from the legal perspective and develop it’s

own basic concepts and tools. On the other hand,

Artificial intelligence has found a lot of uses in

legal world. There is work been done in the field of

creation of better LawBots. These LawBots are

basically robots which can be hired to do the simple

and repetitive legal procedures. For say, filing of a

case by the clerks takes a lot of work but the

procedure is almost the same. Such jobs can be

done by the LawBots. But due to the opposition to

AI killing the jobs, the topic did not develop well.

There is another use of Artificial Intelligence in

Law. It is that of Legal Research.

Before going anywhere now, we define Artificial

Intelligence. Artificial Intelligence are models

targeted at thinking, perception and action. These

models are supported by representations which are

exposed to constraints. These constraints are

enabled by certain algorithms written by computer

scientists.

Now, Legal Research is the process of identifying

and retrieving worthy data which shall be used in

legal problem solving and decision making. The

entire work shall be done by an AI by the same

simple procedures which Lawyers tend to do. The

result would be a better search from the provided

database as well as a better viewpoint of the non-

human intelligence. This shall also trigger new

ideas in the lawyers to pursue the case in a more

wider ambit.

II. LEGAL RESEARCH

The Legal Research is a very simple procedure and

is known to every lawyer practicing law in the court

of law. When a case is presented in the court f law,

the counsel advocate present their viewpoints to

satisfy the judge and lure him into giving the

judgment in his favor. The background of this act is

a daily research work collecting strong arguments

backed by precedents or case laws. Most cases are

such which happen to be similar to some of the

previous, or a part of them is similar to some

previous cases. These arguments presented before

the court hold water only when they are backed by

such case laws or precedents. These chunks of data

are Legal Sources. There are a variety of legal

sources. They are:

1. Legislation

2. Judicial Precedents.

3. Customs

Legislation involves all the laws which are created

and introduced by the ruling government. It is the

most effective and has the ultimate power to win a

case.

Judicial Precedents are case laws, i.e. the judgments

delivered by the courts previously which have

binding power in the current court of law.

Customs are traditions which haven’t been codified

but are practiced by the people in any state. The

custom shall have been practiced from times

immemorial.

For research purpose, the sources are classified into

two. They are:

1. Primary Sources

2. Secondary Sources

Primary Sources are those sources which are

sanctioned by government entity. These government

entities are courts, legislature or executive agencies.

These sources are Case Laws, judgments, Laws and

constitution.

Secondary sources are like a restatement of the

primary law with or without a commentary and a

viewpoint or a suggestion. These secondary sources

are :

1.Legal Dictionaries

2. Words and Phrases

3. Legal Encyclopedias.

4. Annotated Law Reports

5. Legal Periodicals

6. Legal Directories

7. Constitutional Debates

8. News

Such sources are also called Secondary sources.

III. CONCEPTS OF ARTIFICIAL INTELLIGENCE

The concepts in Artificial Intelligence that can be

used for legal research are :

1. Rule Based Expert Systems

2. Search Methods

3. Goal Trees

Rule based expert systems are basically novice

systems. They are also called Deduction System.

We give some pre-recognized rules and according

to them, the system finds out the correct and the

most accurate data required as per the case.

These rules are basic deducing rules which happen

to work for almost every general case. Say, the

lawyer asks for a certain type of case. The first

process is to identify the case. This will need

grasping it’s basic characterstics. Here is where

Rule Based Expert systems come into play. There

will be rules for every type of law to be identified in

a case through simulations or keywords.

These systems were developed in mid 1980s.

Knowledge is basically encapsculated in the form of

simple rules. These rules make a complex case easy

to comprehend and hence, the identification of the

case is done.

There are two approaches to making a expert

system. These approaches are called Forward

chaining and Backward chaining. Forward chaining

rule based expert system is a system where it works

forward from facts to the conclusion. The backward

chaining method is where the system assumes a

hypothesis. It works backward from the hypothesis

to the facts. Whichever facts meet the given ones,

that hypothesis becomes the conclusion. Both

systems are Deduction Systems.

The next step is to answer the questions on it’s own

behavior. This requires it to draw Goal Trees. Goal

Trees are such trees composed of nodes and links

without any loops. They are focused on getting to a

goal. This is one of the most basic concepts in

Artificial Intelligence. Whatever the problem may it

be, it can be solved by drawing a goal tree for the

problem. If we go back the links that are formed, we

would be able to see that the program answers

questions based on it’s own behavior. For say, at the

end of the research done by the AI, if the lawyer

seeks to learn what keyword or question made a

particular document to be in the results, the artificial

Intelligence can straightaway direct it to the specific

keyword, thus, making the research more user

friendly.

The last topic to discuss is Search methods. There

are a variety of search methods in computer science.

There are hill climbing, depth-first, breadth-first,

beam search, optimal and many more.

The intelligence shall be provided with a large

database of cases, laws, journals and all the sources

which were mentioned in the previous section.

Through them, the search methods enable to find

the finest documents for the research work. Also,

the Search methods can be used in conjunction with

Goal trees. This will help understand the user that

why did the system reject certain cases and select

certain cases. Goal trees will be drawn on the

searching pattern. The rules for the search basically

constitute the types of search.

IV. ROSS INTELLIGENCE

Ross Intelligence is one of the finest artificial

intelligence in the field of legal research introduced

to the world. The intelligence is based on concepts

of artificial intelligence like planning, goal trees,

automated search methods from databases etc.

Ross intelligence is known for it’s precise

highlighting fo topics, intuitive queries and law

monitoring. It can monitor the latest legal

developments.

Currently, ROSS operates in three legal areas.

1. Bankruptcy Law.

2. Intellectual Property Law.

3. Labor and Employment Law.

Ross Intelligence is a direct application of the

procedure stated above. The processing of

extraction of information from the large database of

legal material and then, ability to answer questions

on it’s own research which is called behavior, is

done by making goal trees.

Today, there are a number of firms that have

multiplied their success by taking the help of ROSS

intelligence. ROSS intelligence shall further be

developed and learning methods should be made

applicable to it. There is not much data available on

ROSS intelligence.

V. OTHER DIMENSIONS OF LAW

Artificial Intelligence is the new emerging

technology in the modern age. This era is just the

initiation era for artificial intelligence. A non-

human intelligence can be more than a novice

system. It can contribute to international relations

between the countries. Countries today, have

bilateral treaties. An Artificial Intelligence can

promote peace in the world by drafting certain

model multilateral treaties so that all the countries

may benefit. Artificial Intelligence, can perceive

neutrally for all countries, and may help out on

sensitive issues such as Climate change. Thus, there

is more to Artificial Intelligence than it’s use in

Legal Research.

Author’s note

The author/researcher would like to thank Amity University

and Cairia 2018 for giving him this chance to present his idea

via this paper in front of you.

He would also like to thank his friends Prafulla Sahu and

Abhivardhan for their constant support and love.

The author would at last like to thank God.

References

[1] Patrick Henry Winston, Artificial Intelligence, 3rd

edition. [2] https://ocw.mit.edu/courses/electrical-engineering-and-

computer-science/6-034-artificial-intelligence-fall-2010/

[3] V.D. Mahajan,Jurisprudence and Legal Theory.

[4] https://rossintelligence.com/

Role of Artificial Intelligence in Healthcare

Devanshu Srivastava

AMITY Institute of Information Technology (AIIT)

AMITY University

Lucknow, India

E-mail: Devanshucool007@gmail.com

Abstract

Artificial Intelligence tends to duplicate the human

responsive system. The use of artificial intelligence

has increased very rapidly from last 5-10 years and

has now become the most used technology across

the globe. The artificial intelligence has established

itself in almost every field like education, business,

healthcare etc. The AI has proved its importance in

the field of healthcare by analyzing the data and

giving the exact cause of the disease along with its

solutions. AI uses some of the methods like

Machine Learning (ML), Natural Language

Processing (NLP), Neural Networks and many more

for evaluating the disease and giving the best

solutions for curing the disease. The most important

areas where the AI tools are used are Cancer,

Diabetes, Stroke, Neurology and Cardiology.

Key Words: Neurology, Cardiology, Neural

Networks, Natural Language Processing and

Machine Learning.

Introduction

Nowadays, AI techniques have sent a drastic

intuition’s across the healthcare sector and also

pumping the active debate about that whether the AI

technique will replace the human doctors in the

coming future. But according to a study the human

doctors cannot be replaced by the AI doctors in the

near future but instead they can help the physician

in making the better and effective clinical decisions.

The AI technique can help the doctors to study the

disease the cause and its symptoms so that they can

take effective decisions on how to cure that disease

in minimal time. AI analyzes the healthcare data

and big data to collect information’s about the type

of disease and its cause of occurrence in the human

being which helps to solve the problems. AI provide

the realistic data of the problem or the disease by

taking the previous data into account and on that

data it predicts the outcomes and the causes which

can be helpful in saving someone’s life.

The aspects of applying AI in the healthcare are as

follows:

1. Reasons for applying artificial intelligence

in the healthcare sector.

2. The types of data that has to be analyzed.

3. Devices which enables to generate

meaningful AI data.

4. Types of disease that AI has to tackle.

Reasons

The main reason to use AI in healthcare is its

repeatability and reliability. Once the AI is

configured and tested then it can analyze data

continuously without any stress and fatigue and can

generate the effective results. It can be updated with

the clinical data and assumptions to make the result

more beneficial and useable for the human doctors.

It also has the capability of self-learning and auto

correcting without any involvement of the human

brain. AI can help to reduce the errors of diagnostic

and neural tests. It has the capability of self-

evaluating and testing so that it can produce the

desired result and also it is not biased for anything.

So, the chances of errors are very less and almost

negligible. It uses a large patient data to assist

making real-time decisions for health risk and health outcome predictions.

Figure-1 Data types used in artificial intelligence. [1]

HealthCare Data-Types

Before implementing the AI systems in the real

world it has to be tested over and trained over large

set of clinical data of diagnosis, stroke, and diabetes

etc. so that it can generate the genuine results about

the data studied. The data types can vary from

disease to disease and the AI should be capable of

reading all those data and produce different results

for the different type of data. The above figure-1 [1]

shows the different types of data that are being

studied by the AI to produce the results of different

diseases.

AI Devices

AI devices are mainly categorized in two main

sections. The first one includes Machine Learning

(ML) which analyzes the structured data or the data

is which is available online or in any literature form

like images, texts etc.

The second category includes natural Language

Processing (NLP) methods which includes

unstructured data such as clinical notes, medical

prescriptions or data etc. NLP aims to convert the

text into the ML so that AI can easily understand it

and evaluate it. The Figure-2 [2] describes the

architecture of AI in healthcare using NLP data and

ML data.

Types of Disease

Instead of having rich AI content in medical field,

the research mainly focuses on some common

disease types like cancer, neural disease and

cardiovascular disease (figure-3) [3]. Some of the

examples are given below:

1. Cancer: Cancer has been the most

dangerous and major cause of the deaths in

Asia, North America and some parts of

Middle East. There are different types of

cancers like blood cancer, skin cancer, brain

cancer and etc. AI can be used to gather

information’s about different types of cancer

and there reasons for occurrences in a

human being and from that data doctors can

easily focus on the cause of a particular

cancer so that a proper vaccinations can be

made.

2. Neural Disease: Neural disease includes

damage of nerves and neurons in a human

body which carries signal from brain to

other parts of the body. When a particular

nerve is damaged then it is unable to carry

the brain signals which lead to the

disfunctioning of a particular body parts like

hands, legs, finger movement etc. AI helps

to recognize the major causes of neurons

dysfunction by analyzing the medical data of

several patients.

Figure-2: Architecture of AI in healthcare using NLP data and ML analysis. [2]

3. Cardiology: Cardiology is a part of

medicine which deals with the heart diseases

and different parts of the circulatory

systems. This includes Cardiac arrest,

clotting of blood in the heart vessels etc. It is

one of the major disease found in the age

group of 30 above people many people lose

their lives due to heart diseases every year.

The above three diseases are the main focus

because they are very dangerous and takes

many lives every year. AI in these fields can

help to improve the medical facility because

it will predict the exact outcomes and

reasons of the disease.

Figure-3: The leading 10 diseases. [3]

Applications of AI in Stroke

Stroke/Heart Attack is a very common and frequent

occurring disease which causes more than 300

million people to lose their lives worldwide.

Therefore, research on how to prevent stroke is very

important. In the past few years AI techniques have

been used to gather more information’s about heart

related issues. Stroke for 65% of the time, is caused

due to clotting of blood in the heart vessel.

There are many reasons which affect the stroke and

disease mortality. As compared to the previous

methods ML methods are more accurate and precise

in improving prediction performance about the

stroke.

ML and NLP can be used to analyze stroke data of

different patient around the globe and based on that

analysis it can predict the main cause of stroke in a

human being and what age it occurs the most.

Conclusion

I reviewed the reasons of using AI in healthcare and

also presented the various data of different diseases

which are the major cause of death nowadays. After

that I have also discussed various techniques used to

analyze data and give predictions out of those data

Of various patients across the globe which will be

helpful in finding the reasons of the diseases. The

Various techniques that are used in this are Machine

Learning (ML) and Natural Language Processing

(NLP). ML is a structured data which works on the

online or printed data to give the predictions

whereas NLP uses clinical data to give the same.

By using Artificial Intelligence we can save many

lives which we loses due to the lack of knowledge

and decision making system. AI helps to give the

perfect data to the doctors so that they can easily

take decision that what steps to be followed in the

particular disease so that it can be cured.

In the near future most of the health related issue

will be tackled by the AI itself but under the human

supervision.

References

1. (2018)- Data types used in artificial

intelligence. [online]Available.

https://www.google.co.in/search?q=ai+i

n+healthcare+data&source=lnms&tbm=i

sch&sa=X&#imgrc=O9rnblpYjyQIyM:

2. (2018)- Architecture of AI in

healthcare using NLP data and ML

analysis.[online]Available.

https://www.google.co.in/search?biw=13

66&bih=662&tbm=isch&sa=1&ei=9jX

CWrekEsPVvgSF45OYCA&q=roadmap

+of+ai+using+nlp+and+ml&oq=roadma

p+of+ai+using+nlp+and+ml&gs_l=psy-

ab.3...72919.80281.0.80574.18.14.1.0.0.

0.655.3103.0j1j7j5-

2.10.0....0...1c.1.64.psy-

ab..8.0.0....0.nEraZV8ek7w#imgrc=f3sh

wwEEvpDgLM:

3. (2018)- The leading 10 diseases.

[online]Available.

https://www.google.co.in/search?q=ai+i

n+healthcare+data&source=lnms&tbm=i

sch&sa=X&#imgdii=XAB2y81qEij0iM:

&imgrc=O9rnblpYjyQIyM:

Conference Chair

Dr. Pankaj K. Goswami

Assistant Professor, AIIT

Mob. : 9453434364, pkgoswami@amity.edu

Publication Chair

Dr. Parul Verma

Assistant Professor, AIIT

Mob. : 9839289870, pverma1@lko.amity.edu

ISBN 978-93-5300-754-6

CAIRIA-2018