UNIVERSITY AT BUFFALO - STATE UNIVERSITY OF NEW YORK
The Department of Computer Science & Engineering

CSE 642: TECHNIQUES OF ARTIFICIAL INTELLIGENCE - Spring 2001

CSE 642: TECHNIQUES OF ARTIFICIAL INTELLIGENCE:
Computational Theories of Contextual Vocabulary Acquisition

SYLLABUS

Last Modified: 8 February 2001

material is highlighted

---

Index:	Other Relevant Links:
Topics	CSE 642 homepage
Prerequisites	Directory of Documents
Staff
Class Meetings
Texts
Important Dates & Tentative Schedule
Reading
Course Requirements
How to Study
Grading
Incompletes
Academic Integrity

---

As part of an interdisciplinary project to develop a computational cognitive model of a reader of narrative text, we are developing a computational theory of how natural-language-understanding systems can automatically acquire new vocabulary by determining from context the meaning of words that are unknown, misunderstood, or used in a new sense. 'Context' includes surrounding text, grammatical information, and background knowledge, but no external sources. Our thesis is that the meaning of such a word can be determined from context, can be revised upon further encounters with the word, "converges" to a dictionary-like definition if enough context has been provided and there have been enough exposures to the word, and eventually "settles down" to a "steady state" that is always subject to revision upon further encounters with the word. The system is being implemented in the SNePS knowledge-representation and reasoning system.

During the course, we will be reading through as much as possible of the large literature in computational linguistics, reading education, and second-language (L2) learning that is concerned with what is variously called "vocabulary acquisition from context", "vocabulary acquisition in context", "incidental vocabulary acquisition", "word-sense disambiguation", etc.

PREREQUISITE:

However, students from all of the cognitive sciences and from education are welcome to enroll. No programming background is necessary.

Professor:

Dr. William J. Rapaport, 214 Bell Hall, 645-3180 x 112, rapaport@cse.buffalo.edu
Office Hours: Mon & Fri 1:00 - 1:50 p.m.; and by appointment.

1. Rapaport, William J., & Ehrlich, Karen (2000), "A Computational Theory of Vocabulary Acquisition", in Lucja M. Iwanska, & Stuart C. Shapiro (eds.), Natural Language Processing and Knowledge Representation: Language for Knowledge and Knowledge for Language (Menlo Park, CA/Cambridge, MA: AAAI Press/MIT Press): 347-375.
   • Errata: (postscript) (PDF)

2. Rapaport, William J., & Kibby, Michael W. (2000), "Contextual Vocabulary Acquisition: Development of a Computational Theory and Educational Curriculum" (NSF grant proposal).

3. Rapaport, William J. (2001), "A (Partial) Bibliography (in Chronological Order) of (Computational) Theories of Contextual Vocabulary Acquisition".

IMPORTANT DATES & TENTATIVE SCHEDULE:

DAYS	DATES	TOPICS
	Jan 16 - Jan 18	Intro to computational theories of contextual vocabulary acquisition
	Jan 23 - rest of semester	student and faculty reports on the literature
Wed	Jan 31	Center for Cognitive Science colloquium: J.-P. Koenig, UB Dept. of LIN, "What's 'In' a Word?" 2:00 - 3:30 p.m., 280 Park
Wed	Feb 14	Center for Cognitive Science colloquium: Carl Alphonce, UB Dept. of CSE, "Computational Implementation of Theories" 2:00 - 3:30 p.m., 280 Park
Thu	Feb 15	CSE Grad Conference CLASS CANCELLED
Sat-Sun	Mar 3 - Mar 13	Spring Break (no classes)
Tue	Mar 13	*** TERM-PROJECT PROPOSALS DUE ***
Fri	Mar 16	*** Last day to withdraw with a grade of `R' ***
Thu	Apr 26	Last Class: Summary
Thu	May 3	*** TERM PROJECTS DUE ***

1. THE COURSE WILL BE RUN AS A SEMINAR. Thus, you will be expected to read and give oral presentations on several papers from the online bibliography (or elsewhere), as mutually agreed upon. You are strongly encouraged to read the papers that other students will be presenting.

2. See "How to Read (a Computer Science Text)".

1. You should subscribe to, and regularly monitor, the newsgroup sunyab.cse.642. You may post questions and comments there that are of general interest to the entire class. From time to time, information about readings, etc., will be posted to the newsgroup. This newsgroup will be archived in the CSE 642 Newsgroup Archive.

2. There are two main requirements:

   • You must attend virtually all classes, actively participate in class discussions, and be prepared to give your oral presentations on time.

   • There will be a term project. This will be a 10-15-page, "typed", double-spaced report (which may, but need not, include an appendix detailing a computational implementation).

     The final report for an implementation project must include an appendix (beyond the 10-15-page report itself) with annotated sample runs and the commented program code.

     The report might be:

     • new tests of our current contextual-vocabulary-acquisition system
     • an extension to the grammar
     • a determination of which standard learning theory we currently use
     • an investigation of other learning theories that might be relevant
     • coding the background knowledge for an appropriate text
     • a (partial) literature review of:
       • computational theories of contextual vocabulary acquisition
       • L1 theories of contextual vocabulary acquisition
       • L2 theories of contextual vocabulary acquisition
       • theories of context
     • or any other mutually-agreed-upon topic.

     If several of you choose to work together on a team project, your report must include a statement clearly outlining each person's contribution.

     Your topic must be approved by me in advance, and a proposal (perhaps in the form of an extended summary) of 6-8 "typed", double-spaced pages or--even better--a first draft must be turned in no later than Tuesday, March 13.

     The term project is due no later than Thursday, May 3 (= the first day of final exams). Late projects will be accepted at a full-letter-grade-per-day discount (e.g., an A project, 1 day late, gets a B, not an A-).

3. Students should notify Prof. Rapaport within the first two weeks of class if they have a disability which would make it difficult to carry out course work as outlined (requiring note-takers, readers, extended test time).