cs626-449: nlp, speech and web-topics-in-ai

20
CS626-449: NLP, Speech and Web- Topics-in-AI Pushpak Bhattacharyya CSE Dept., IIT Bombay Lecture 35: Semantic Relations; UNL; Towards Dependency Parsing

Upload: meryl

Post on 07-Jan-2016

20 views

Category:

Documents


0 download

DESCRIPTION

CS626-449: NLP, Speech and Web-Topics-in-AI. Pushpak Bhattacharyya CSE Dept., IIT Bombay Lecture 35: Semantic Relations; UNL; Towards Dependency Parsing. Web at a glance. Google indexes more than 8 billion pages Dominated by English Large part of world is deprived of this knowledge. - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: CS626-449: NLP, Speech and Web-Topics-in-AI

CS626-449: NLP, Speech and Web-Topics-in-AI

Pushpak BhattacharyyaCSE Dept., IIT Bombay

Lecture 35: Semantic Relations; UNL; Towards Dependency Parsing

Page 2: CS626-449: NLP, Speech and Web-Topics-in-AI

2

Web at a glance

Google indexes more than 8 billion pages

Dominated by English Large part of world is deprived of

this knowledge

Page 3: CS626-449: NLP, Speech and Web-Topics-in-AI

3

Search Engines Today

Keyword based Irrelevant results

Meaning not taken into account

Language specific No search possible across language No translation possible

Page 4: CS626-449: NLP, Speech and Web-Topics-in-AI

4

Future of the World Wide Web

WWW

User

Translationinterface

User

User

User

Translationinterface

Page 5: CS626-449: NLP, Speech and Web-Topics-in-AI

5

Features

Meaning based More relevant results

Multilingual Query in English Fetch document in Hindi Show it in English

Page 6: CS626-449: NLP, Speech and Web-Topics-in-AI

6

Machine Translation

Translate from one language to other

Two approaches Direct

One step Using Intermediate language

Two step

Page 7: CS626-449: NLP, Speech and Web-Topics-in-AI

7

Interlingua Interlingua

Intermediate language for machine translation

Step one Convert from source language text to

interlingua Step two

Produce target langauge text from interlingua

UNL : an interlingua in UNL system

Page 8: CS626-449: NLP, Speech and Web-Topics-in-AI

8

Internet for the Masses

Internet

Spanishinterface

English viewer

Hindi viewer

English interface

Spanishviewer

Hindi interfac

e

Page 9: CS626-449: NLP, Speech and Web-Topics-in-AI

9

A Semantic Graph

in: modifiera: indefinite

the: definite

student

past tense

agent

bought

objecttime

computer

new

June

modifier

The student bought a new computer in June.

Page 10: CS626-449: NLP, Speech and Web-Topics-in-AI

10

UNL representation

Ram is reading the newspaper

Representation of Knowledge

Page 11: CS626-449: NLP, Speech and Web-Topics-in-AI

11

Knowledge Representation

Ram

read

newspaper

agt obj

UNL Graph - relations

Page 12: CS626-449: NLP, Speech and Web-Topics-in-AI

12

Knowledge Representation

Ram(iof>person)

read(icl>interpret)

newspaper(icl>print_media)

UNL Graph - UWs

agt obj

Page 13: CS626-449: NLP, Speech and Web-Topics-in-AI

13

Knowledge Representation

Ram(iof>person)

read(icl>interpret)

newspaper(icl>print_media)

@entry@present@progress

@def

Ram is reading the newspaper

UNL graph - attributes

agt obj

Page 14: CS626-449: NLP, Speech and Web-Topics-in-AI

14

The boy who works here went to school

plt

agt@ entry @ past

school(icl>institution)

go(icl>move)

boy(icl>person)

work(icl>do)

here

@ entry

agt plc

:01

Another Example

Page 15: CS626-449: NLP, Speech and Web-Topics-in-AI

15

UNL System

Page 16: CS626-449: NLP, Speech and Web-Topics-in-AI

16

The World-wide Universal Networking Language (UNL) Project

UNL

English Russian

Japanese

Hindi

Spanish

Language independent meaning representation.

Marathi

Page 17: CS626-449: NLP, Speech and Web-Topics-in-AI

17

The UNL System: An Overview

Page 18: CS626-449: NLP, Speech and Web-Topics-in-AI

18

Universal Networking Language

Universal Words (UWs) Relations Attributes Knowledge Base

Page 19: CS626-449: NLP, Speech and Web-Topics-in-AI

19

UNL Graph

obj

agt

@ entry @ past

minister(icl>person)

forward(icl>send)

mail(icl>collection)

He(icl>person)

@def

@def

gol

He forwarded the mail to the minister.

Page 20: CS626-449: NLP, Speech and Web-Topics-in-AI

20

UNL Expression

agt (forward(icl>send).@ entry @ past, he(icl>person))

obj (forward(icl>send).@ entry @ past, minister(icl>person))

gol (forward(icl>send ).@ entry @ past, mail(icl>collection). @def)