LING 388 Language and Computers - PowerPoint PPT Presentation

Loading...

PPT – LING 388 Language and Computers PowerPoint presentation | free to download - id: a5bdc-OTc4M



Loading


The Adobe Flash plugin is needed to view this content

Get the plugin now

View by Category
About This Presentation
Title:

LING 388 Language and Computers

Description:

Tuesday/Thursday 11:00 am-12:15 pm. Office Hours: (TBA) After class. TA: ... AI-Complete problem (Killer App, 'Holy Grail') Machine Translation: Architecture ... – PowerPoint PPT presentation

Number of Views:31
Avg rating:3.0/5.0
Slides: 24
Provided by: sandiw
Category:

less

Write a Comment
User Comments (0)
Transcript and Presenter's Notes

Title: LING 388 Language and Computers


1
LING 388 Language and Computers
  • Lecture 1
  • 9/02/03
  • Sandiway FONG

2
Administrivia
  • Where?
  • Franklin 210
  • When?
  • Tuesday/Thursday 1100 am-1215 pm
  • Office Hours
  • (TBA) After class
  • TA
  • Charles Lin (clin_at_email.arizona.edu)
  • Email
  • sandiway_at_cs.arizona.edu
  • sandiwayfong_at_yahoo.com (temporary)

3
Administrivia
  • Lecture slides
  • Available on the web in Adobe PDF format after
    each class
  • Temporary location
  • http//external.nj.nec.com/homepages/sandiway/ling
    388/Lecture1.pdf
  • You need to go to class to understand the lecture
    slides
  • Programming
  • Some homeworks will involve simple programming
    exercises such as writing grammars in logic
  • Access to PCs
  • Do not panic we will help

4
Adminstrivia
  • Textbooks
  • None formally required
  • All necessary background will be introduced
    during the lectures
  • However a PROLOG (Programming in Logic) textbook
    is recommended

5
Adminstrivia
  • Homework Assignments
  • There will be 6-8 homeworks and a final
    project/paper
  • You have one week to turn in the homework
  • Do not fall behind!
  • Feedback on how well youre understanding the
    material
  • If you cant do the assignment, get help
    immediately!
  • You can talk about the assignments, but you must
    write them up independently
  • Homeworks involving writing grammars
  • There will be some flexibility wrt the turn-in
    date here for those who have no exposure to
    programming
  • But you must ask for help immediately!

6
Language and Computers
  • Goals of the course
  • Introduction to natural language processing aka
    computational linguistics
  • Good idea of what computers can do
  • Good idea of what is hard to do
  • Be able to write simple formal grammars
  • Use logic as a programming language

7
The Field of Computational Linguistics
  • Concerned with
  • Computational methods for working with natural
    languages
  • Intersects with
  • Computer science (CS)
  • Mathematics/Statistics
  • Artificial intelligence (AI)
  • Linguistic Theory
  • Psycholinguistics
  • the human sentence processor

8
Properties of Natural Languages
  • Grammar
  • portions innate/learnt
  • Lexicon
  • specific properties of words
  • Example
  • Innate knowledge
  • Which report did you file without reading?
  • Example
  • Ambiguity
  • Where can I see the bus stop?

9
Properties of Natural Languages
  • Ungrammaticality
  • Example
  • Which book did you file the report without
    reading?
  • Contrast with
  • Which report did you file without reading?

10
Properties of Natural Languages
  • Interpretation
  • Examples
  • John is too stubborn to talk to
  • John is too stubborn to talk to Bill

11
The Human Sentence Processor
  • Properties
  • Garden-pathing
  • Ian told the man that he hired a story
  • Ian told the man that he hired a secretary
  • Non-uniform processing cost
  • The reporter who the senator attacked admitted
    the error
  • The reporter who attacked the senator admitted
    the error
  • Goal to mimic human performance?

12
Sentence Analysis
  • Representation?
  • Syntax tree

13
Sentence Analysis
  • Logic representation of events
  • Examples
  • John is too stubborn to talk to
  • event(e) agent(x,e) theme(John,e)
  • John is too stubborn to talk to Mary
  • event(e) agent(John,e) theme(Mary,e)

14
Computer Science/Artificial Intelligence
  • Areas
  • Formal Grammars
  • Algorithms
  • Complexity Theory
  • Knowledge Representation
  • Search
  • Heuristics
  • Statistical Methods

15
Many Applications…
  • Range from Lightweight to AI-Complete
  • Spell/Grammar Checkers
  • Cellphone Text Entry
  • Information Retrieval (IR)/Question-Answering
    (QA) Systems
  • Speech Recognition/Synthesis
  • Sentence/Dialogue Understanding
  • Document Summarization
  • Machine Translation (MT)

16
Some Examples…
17
Spell Checking
  • Word processing
  • Information Retrieval (IR)

18
T9 Cellphone Text Entry
  • T9-enabled text entry system on Nokia phones
  • How to enter words on a numeric keypad quickly?
  • Reference
  • http//www.t9.com/

19
T9 Cellphone Text Entry
  • Why is it useful?
  • Address/contact information entry
  • Text messaging
  • SMS big thing in Europe
  • How does it works?
  • Stored database of common English words,
    including proper names
  • Predicts the possible words for sequences of key
    presses

20
Speech Recognition/Synthesis
  • Real-time application
  • Speaker dependent/independent
  • Controlled/unrestricted vocabulary
  • Continuous/isolated speech
  • Examples of Applications
  • Hands-free environment Automobiles
  • Disabilities Dictation/Reading
  • Directory Inquiries ATT Toll-Free Directory
  • Speech-to-speech MT

21
Machine Translation
  • Text-to-text, speech-to-speech,…
  • AI-Complete problem (Killer App, Holy Grail)

22
Machine Translation Architecture
  • Whats inside that black box?
  • Grammars and dictionaries for two languages
  • Points of translation between the two languages
  • What are the possible architectures?

??
English
Japanese
23
Demo…
  • PDA prototype
  • PC-based version
  • Reference
  • http//www.nec.co.jp/press/en/0301/2202.html
About PowerShow.com