deep learning for text analytics

Post on 14-Jan-2017

347 Views

Category:

Technology

2 Downloads

Preview:

Click to see full reader

TRANSCRIPT

(Technical) Big Data Analytics

for non-technical end-users

Erik Tromp – CEO UnderstandLing

Contents• (Short) into• Rationale• Tuktu platform• Deep learning for computational linguistics• CEMistry – Customer Experience Monitoring on steroids

(Short) Intro

• Big data science experts• Specialisms• Computational Linguistics• Customer Experience Management

• Service: strategic advices all the way to operational implementation• Own platform: Tuktu• Soon: own product: CEMistry• Trainings/education on big data science

(Short) IntroQuantify every touchpoint of a customer with your company

4 major areas

• Text Analytics• Web Analytics• Mobile Analytics

• CRM/Backend Analytics

(Short) Intro• Erik Tromp• Age: 28• CEO UnderstandLing• Graduated on Sentiment Analysis in 2011• Multilingual Sentiment Analysis on Social Media

• Software engineer – Scala• Machine learning• Author of platform Tuktu

Contents• (Short) into• Rationale• Tuktu platform• Deep learning for computational linguistics• CEMistry – Customer Experience Monitoring on steroids

Rationale

Big data science allows to utilize opportunities

Rationale

Big data science allows to utilize opportunities

Big data science drives business

Rationale

Big data science allows to utilize opportunities

Big data science drives business

But is very much a technical revolution, with business implications

Rationale

Many companies want to utilize the opportunities big data science brings

Rationale

Many companies want to utilize the opportunities big data science brings

These companies do not have sufficient capabilities to do so

Rationale

Many companies want to utilize the opportunities big data science brings

These companies do not have sufficient capabilities to do so

Nor are there many suppliers that can do tech, analytics and know their business

Rationale

But these companies often do have their own (business) analysts

Rationale

IDEA

Make big data science accessible to non-technical users

Contents• (Short) into• Rationale• Tuktu platform• Deep learning for computational linguistics• CEMistry – Customer Experience Monitoring on steroids

Tuktu

http://www.tuktu.io

https://github.com/UnderstandLingBV/Tuktu

Tuktu – Early Days• Started off as a personal project to make life easier• Out of a collaboration with the Maastricht University• Idea: save time on coding/engineering, focus on logic and

functionalities

Tuktu – Early Days• Started off as a personal project to make life easier• Out of a collaboration with the Maastricht University• Idea: save time on coding/engineering, focus on logic and

functionalities

Instead of writing code over and over again, have it present and configure its building blocks

Tuktu – Early Days• Started off as a personal project to make life easier• Out of a collaboration with the Maastricht University• Idea: save time on coding/engineering, focus on logic and

functionalities

Instead of writing code over and over again, have it present and configure its building blocks

In a visual and straightforward way!

Tuktu – Now

Your one-stop shop for everything big data science

Tuktu – Now• Realtime and batch

processing• Synchronous and

asynchronous processing• REST API• Drag-and-drop modelling of

jobs• Distributed file system: TDFS• Key/value-sture: TuktuDB• Real-time visualization

• Web analytics support• Scheduling• No master/slave architecture• Local or distributed

computing• Machine learning• Deep learning• Cross-platform due to JVM• Easy installation: just unzip!

Tuktu

DEMO

Contents• (Short) into• Rationale• Tuktu platform• Deep learning for computational linguistics• CEMistry – Customer Experience Monitoring on steroids

Deep Learning for Computational Linguistics

IDEA

Learn language models generically

Deep Learning for Computational Linguistics

IDEA

Learn language models generically

Model every CL-problem on top on the generic model

Deep Learning for Computational Linguistics

This way, we can do almost any task on almost any language

Without too much/with less effort

Deep Learning for Computational Linguistics

How?

Deep Learning for Computational Linguistics

There are many linguistics resources available

Sadly; most is for EnglishIn particular: Annotated Treebanks for deep parsing

Deep Learning for Computational Linguistics

We can use this however

Deep Learning for Computational Linguistics1.Co-train word vectors for target language and English

Deep Learning for Computational Linguistics1.Co-train word vectors for target language and English2.Train parsing models on English language

Deep Learning for Computational Linguistics1.Co-train word vectors for target language and English2.Train parsing models on English language3.Co-finetune models on co-trained word vectors

Deep Learning for Computational Linguistics1.Co-train word vectors for target language and English2.Train parsing models on English language3.Co-finetune models on co-trained word vectors4.Pre-train (recursive) auto-encoder using parsing model for

target language

Deep Learning for Computational Linguistics1.Co-train word vectors for target language and English2.Train parsing models on English language3.Co-finetune models on co-trained word vectors4.Pre-train (recursive) auto-encoder using parsing model for

target language5.Use recursive auto-encoder for specific task in target

language• Topic detection, sentiment analysis, named entity recoginition,

authorship profiling

Deep Learning for Computational Linguistics

DEMO

Unsupervised parsing in Dutch

Contents• (Short) into• Rationale• Tuktu platform• Deep learning for computational linguistics• CEMistry – Customer Experience Monitoring on steroids

CEMistryTEXT ANALYTICS

MOBILE ANALYTICS

WEB ANALYTICS

BACKEND/CRM

Customer Profile

Tuktu.jsVisitor Customer

EventsPage views

Link

“SDK”Visitor Customer

EventsApp Triggers

Link

User CustomerCollectors

EventsCommunicati

on(NLP)

Link

(Database)ConnectorsCustomer

EventsTransactions

Link

Questions?

Erik TrompCEO UnderstandLing

erik@understandling.com

http://www.understandling.comhttp://www.tuktu.io

http://www.linkedin.com/in/eriktromphttps://github.com/UnderstandLingBV/Tuktu

Talk to us on Gitter! https://gitter.im/UnderstandLingBV/Tuktu

top related