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

CSE 663: ADVANCED TOPICS IN KNOWLEDGE REPRESENTATION Spring 2002

SYLLABUS

Graduate standing and either CSE 563 (Knowledge Representation) or CSE 572 (Knowledge-Based AI) or CSE/LIN 567 (Computational Linguistics); or else permission of instructor.

STAFF:

Professor:

Dr. William J. Rapaport, 214 Bell Hall, 645-3180 x 112, rapaport@cse.buffalo.edu
Office Hours: To be announced, and by appointment.

Knowledge representation (KR), more properly called "knowledge representation and reasoning" (KRR), is the part of AI that is concerned with the techniques for representing and reasoning about the information to be used by an AI program.

• "The core of AI is ... knowledge representation"
  -- Graeme Hirst, Canadian AI #21 (October 1989): 10.

• "The key to finding powerful procedures that can solve [AI] ... problems is to discover appropriate representations of the relevant information. Once a representation is given that `make[s] the right things explicit and expose[s] natural constraints' [Winston], it is a much simpler matter to devise purely computational procedures to manipulate the information, still in its encoded representation."
  -- G. N. Reeke, Jr., & Gerald M. Edelman (1988), "Real Brains and Artificial Intelligence," Daedalus (Winter 1988): 146-147.

• "What structures must be built into a system to allow it to learn? This is a central question for current AI, and the answer depends on issues of knowledge representation: How should knowledge be represented? Out of what components ... are knowledge structures built?"
  -- David Waltz, "The Prospects for Building Truly Intelligent Machines," Daedalus (Winter 1988): 192.

This semester's implementation of CSE 663 will be a practical, research-seminar-style course in applying knowledge-representation and reasoning techniques to two current research projects:

1. Prof. Michael W. Kibby (Department of Learning and Instruction and Center for Literacy and Reading Instruction) and I are principal investigators on a National Science Foundation grant, under the Research on Learning and Education (ROLE) program:
   CONTEXTUAL VOCABULARY ACQUISITION:
   Development of a Computational Theory and Educational Curriculum

   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.

   We propose:

   (a) to extend and develop algorithms for computational contextual vocabulary acquisition (CVA): learning, from context, meanings for "hard" word: nouns (including proper nouns), verbs, adjectives, and adverbs,

   (b) to unify a disparate literature on the topic of CVA from psychology, first- and second-language (L1 and L2) acquisition, and reading science, in order to help develop these algorithms, and

   (c) to use the knowledge gained from the computational CVA system to build and to evaluate the effectiveness of an educational curriculum for enhancing students' abilities to use deliberate (i.e., non-incidental) CVA strategies in their reading of science, math, engineering, and technology texts at the middle-school and college undergraduate levels: teaching methods and guides, materials for teaching and practice, and evaluation instruments.

   The knowledge gained from case studies of students using our CVA techniques will feed back into further development of our computational theory.

   The work involves using natural-language-processing techniques, such as ATN (augmented-transition-network) grammars, and the SNePS knowledge representation and reasoning system.

   For more information on the project, see:

   • 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.

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

   • Rapaport, William J. (2000), "A (Partial) Bibliography (in Chronological Order) of (Computational) Theories of Contextual Vocabulary Acquisition".

2. Prof. Stuart C. Shapiro (Department of Computer Science and Engineering), Dr. Peter Winkelstein (Department of Pediatrics and Children's Hospital), and I are investigating the use of the SNePS knowledge representation and reasoning system, with its natural-language-processing capabilities, for reading, understanding, and making recommendations about hospital discharge summaries.

   The work involves using natural-language-processing techniques such as ATN (augmented-transition-network) grammars, and/or the LKB system, and/or information extraction systems (e.g., TextPro) to parse hospital discharge summaries into a SNePS representation for further processing by SNePS's reasoning facilities, possibly with an eye towards "translating" the discharge summary (intended as a means of communication between medical professionals) into plain English for lay persons (e.g., patients and their families) to be able to understand.

   Further information on the project can be found at: http://www.cse.buffalo.edu/~rapaport/mednlp.ps.

IMPORTANT DATES & TENTATIVE SCHEDULE:

Note: For the record, I have adjusted some of the dates below to reflect what we actually did in class, rather than on what I had hoped to do :-)

DAYS	DATES	DESCRIPTION	ASSIGNMENTS	READINGS
Tue	Jan 22	Introduction to the course and the research projects
	Jan 22-31	Contextual vocabulary acquisition		See "Contextual Vocabulary Acquisition"
Wed	Jan 23	CCS Colloquium David Pierce, UB CSE "Machine Learning Strategies for Corpus-Based Natural-Language Processing" 2:00 - 4:00 p.m., 280 Park
	Feb 5	Medical natural-language processing		See "Medical Natural-Language Processing Project"
	Feb 5-19	SNePS	SNePS Tutorial Martins's KR text, p. 36, #10; due Tue Feb 26	See "SNePS readings" Section on SNePS from Martins text
Thu	Feb 14		Choose CVA or MedNLP project or other approved topic negotiated with me
Tue	Feb 19	UB CSEGSA Grad Conference
Wed	Feb 20	CCS Colloquium Suzanne Stevenson, U/Toronto CS "Learning Semantic Classes of Verbs from Syntactic Frequencies" 2:00 - 4:00 p.m., 280 Park
Thu	Feb 21	SNePS Tutorial due
Tue	Feb 26	Martins HW due
	Feb 26- May 2	Student presentations ("master classes")		Davis, Randall; Shrobe, Howard; & Szolovits, Peter (1993), "What Is a Knowledge Representation?", AI Magazine 14(1): 17-33. Woods, William A. (1975), "What's in a Link: Foundations for Semantic Networks", in Daniel G. Bobrow & Allan Collins (eds.), Representation and Understanding: Studies in Cognitive Science (New York: Academic Press): 35-82. McDermott, Drew (1981), "Artificial Intelligence Meets Natural Stupidity", in John Haugeland (ed.), Mind Design: Philosophy, Psychology, Artificial Intelligence (Cambridge, MA: MIT Press, 1981): 143-160.
Wed	Mar 5	CCS Colloquium Ernest Lepore, Rutgers Center for Cognitive Science "An Abuse of Context in Semantics" 2:00 - 4:00 p.m., 280 Park
Thu	Mar 14	Term-project progress report due.
Fri	Mar 22	Last day to resign with grade of "R"
Sat-Sun	Mar 23-31	Spring Break: no classes
Thu	May 2	Last class
Mon	May 6	Term-project reports due in my office (Bell 214) or mailbox (Bell 211), no later than 5:00 p.m.

Since this is a graduate-level, research-seminar style course, you will be expected to read all assigned readings as well as read relevant literature on your own. For assistance on how to get the most out of your reading, see "How to Read (a Computer Science Text)".

1. You will be expected to attend and participate in all classes. There will be occasional homework assignments, and there will be a term project. If you wish your project to be used as your master's project, please let me know right away, since I would like to schedule extra meetings with such students.

2. Be sure to get a classmate's phone number (for instance, 1 or 2 people sitting next to you in class, whoever they are!) so that you will not miss assignments or announcements in the unlikely event that you miss a class.

3. You should subscribe to, and regularly monitor, the newsgroup sunyab.cse.663. You may post questions and comments there that are of general interest to the entire class. From time to time, announcements and information will be posted to the newsgroup. This newsgroup will be archived in the CSE 663 Newsgroup Archive.

4. Just as you cannot expect to learn how to drive a car by reading about it or by watching other people do it, the same holds true for doing computer science research. Do your work on time--this is one course you simply cannot cram for at the last minute, so don't even try! I cannot stress this strongly enough. The project may be fairly time-consuming, so please consider your other commitments, and plan your time accordingly.

5. 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).

1. The term project for the course is to take a paragraph of text (in most cases, to be provided by me), and to use SNePS to represent the information in the text as well as any background information (including "rules") that would be needed to understand and reason about the text.

2. Project Policies:
   For the final project report, you will be expected to hand in a conference-style paper, typed or printed from a computer file, on 8.5 by 11 inch paper (stapled in the upper left-hand corner, without sprocket holes, and with your own title page). (Please do not use folders or covers, unless your report is too thick to be stapled.) I strongly suggest that you learn to use Latex and ispell.

   Your report should include a syntax and semantics of your knowledge representations, and annotated sample runs of a working SNePS program (including documented code).

   Thus, each report must consist of the following components:

   • descriptive title (not: "Project 1")

   • your name, the course number (CSE 663), and the date.

   • abstract of project (a 1-paragraph summary)

   • description of the project (the body of the paper)

   • list of references (if appropriate)

   For stylistic information and general help on writing, see my Writing Guidelines

   HOW TO STUDY:

   For general advice on how to study for any course, see my web page, "How to Study".

   GRADING:

   Your final course grade will be a weighted average (at least 50-50, but possibly 66 2/3-33 1/3) of (1) your class attendance, class participation, and assignments, and (2) your grade on the project.

   Incompletes:

   It is University policy that a grade of Incomplete is to be given only when a small amount of work or a single exam is missed due to circumstances beyond the student's control, and that student is otherwise doing passing work. I will follow this policy strictly! Thus, you should assume that I will not give incompletes :-) Any incompletes that I might give, in a lapse of judgment :-), will have to be made up by the end of the Fall 2002 semester. Note that this supersedes the more lenient make-up date published in the university catalog! For more information on Incomplete policies, see the web page, "Incompletes".

   ACADEMIC INTEGRITY:

   While it is acceptable to discuss general approaches with your fellow students, the work you turn in must be your own. It is the policy of this department that any violation of academic integrity will result in an F for the course, that all departmental financial support including teaching assistanceship, research assistanceship, or scholarships be terminated, that notification of this action be placed in the student's confidential departmental record, and that the student be permanently ineligible for future departmental financial support. If you have any problems doing the homeworks or projects, consult Prof. Rapaport. Please be sure to read the webpage, "Academic Integrity: Policies and Procedures", which spells out all the details of this, and related, policies.
   
   
   

   ---

   William J. Rapaport (rapaport@cse.buffalo.edu)
   file: 663/S02/syl.09ap02.html