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

CSE 717: SEMINAR: CONTEXTUAL VOCABULARY ACQUISITION Fall 2007

SYLLABUS

• The Project: CONTEXTUAL VOCABULARY ACQUISITION: From Algorithm to Curriculum

  We are (1) developing a computational theory of how natural-language-understanding systems—including both computational and human cognitive agents—can automatically acquire new vocabulary by determining from context a meaning for words that are unknown, misunderstood, or used in a new sense, and (2) adapting the algorithms for doing this to a curriculum so that these methods can be taught to students in a classroom setting.

  We propose:

  1. 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,

  2. 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

  3. 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 Seminar: The seminar will involve:
  1. reading research literature on CVA from computational linguistics, psychology, and education;

  2. using the SNePS knowledge representation and reasoning system to test and/or augment our algorithms; and/or

  3. using natural-language-processing techniques such as ATN (augmented-transition-network) grammars to augment the natural-language front end to our system.

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

IMPORTANT DATES & TENTATIVE SCHEDULE:

Note: As the semester progresses, I will adjust some of the dates below to reflect what we actually do in class, rather than on what I hope to do:-)

1. The first set of readings will be 4 papers on the CVA project and 3 papers on SNePS (see the reading assignments above).

2. The second set of readings will be the "essential CVA readings". I would like you to read all of them. Moreover, for at least 10 of the approximately two dozen readings from this second set, you must submit a 1-paragraph commentary: some insight or idea that struck you from the reading, an extension or application of the ideas, a question that the reading inspired (perhaps informed by some further reading on your part), or any other commentary inspired by the reading. The commentaries should not be summaries of the reading.(*)

   These are to be submitted on-line to the course Listserv on a weekly basis.

   You will also be expected to submit response postings; the commentaries and responses should help our class discussions as well as your own thinking.

   The 10 that you comment on must include the 8 listed at "Really Essential CVA Readings", plus at least 2 more.

   There are about 12 weeks in the semester after the required CVA and SNePS readings; so, you should should read at least 1 per week, starting from the first of those 12 weeks!

These commentaries will be read and recorded, but not graded. However, the number that you do satisfactorily will be factored into your final grade.

HOW TO READ and HOW TO STUDY:

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

• For advice on how to read technical papers in any field (including computer science, psychology, education, etc.), see "How to Read (a Computer Science Text)".

• Your main job besides reading the CVA papers and participating in class discussions will be a computer-based, CVA-related project. See the "Projects" webpage for suggestions.

• At the end of the semester, you will be expected to hand in a conference-style report, typed or printed from a computer file, on 8.5 by 11 inch paper (stapled in the upper left-hand corner, with your own title page [not the priner's cover page!]), describing the project you worked on this semester. (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.

• The intended audience is the person who will be continuing your project next semester, so the report should be self-contained and detailed enough to tell that (possibly hypothetical) person exactly what has been done, why it was done, what problems were faced, how you solved them, what needs to be done next, and how to get started doing it. This might include a well-documented listing of any code you write or well-commented demos or sample runs. It might also include the syntax and semantics of any new SNePS case frames that you use. In any case, any SNePS networks that you include should be fully explained.

  In the real world, you will be expected to write papers, either for presentation at conferences, publication in journals, or presentation to your boss or co-workers. No one reads computer programs except the programmer him- or herself, or someone else who has to modify the program. Users and other people want to read about the program, what it does, how it works, etc., and to see it in action. Consequently, the main product of your work is the paper, not the program! In the paper, you should say what you have done, and say (in English summary, not in programming detail) how you have done it. It should also include annotated examples of your program in action. These should be well chosen to illustrate the range of performance of your program. The examples should not be redundant, nor included merely because they look complicated. Each example should illustrate a particular ability of your program. Nevertheless, the reader will assume that your program does nothing interesting that isn't illustrated! You should read Goldfain, Albert (2003), "Computationally Defining 'harbinger' via Contextual Vocabulary Acquisition" [PDF], to see an example of a such a paper.

  Any program listings should either be presented as figures throughout the paper, or as an appendix. In either case, any listings should be included as documentation for what you say in the paper.

  Thus, each report must consist of the following components:

  • descriptive title (not: "Project 1")

  • your name, the course number (CSE 717), & date completed (e.g., April 30, or whatever).

  • abstract of project (a 1-paragraph summary)
    • Do not assume that your hypothetical reader will have access to the abstract or, if they do have access, that they will read it. Consequently, your paper should include—and probably begin with— any important information from the abstract.

  • description of the project (the body of the paper), including:
    • syntax and semantics of new SNePS case frames,

    • full explanation of any SNePS networks.

  • list of references (if appropriate)
    • Please be sure to cite only published materials (or unpublished papers from websites): Do not cite websites unless there is a very good reason to do so. If in doubt, ask me.

  • annotated sample runs (could be part of the description of the project) and commented code (could be as an appendix), if appropriate.

• I will provide more detailed information as the semester progresses. You should also read my web document, "How to Write", for information on how to prepare the report and for guidance on grammatical points.

ATTENDANCE, HOMEWORKS, ASSIGNMENTS, LISTSERV:

1. You will be expected to:
   1. attend all seminar meetings (attendance will be taken),
   2. participate in class discussions,
   3. present at least one, hopefully two, and possibly three oral progress reports during the seminar meetings, and
   4. complete all readings and assignments on time.

2. Email list:

   You will automatically be placed on an email list (a "Listserv") for the course. If you do not normally read email at the email address that UB has as your official address, please either do so for this course, or else have your mail forwarded. I will use this list as my main means of communicating with you out of class. And you can use it to communicate with the rest of us.

   You may send questions and comments that are of general interest to the entire class using the Listserv: Just send them to:

   cse717-sp07-list@listserv.buffalo.edu

   You can also send email just to me, at:

   rapaport@cse.buffalo.edu

   In any case, be sure to fill in the subject line, beginning with "CSE717: " so that my mailer doesn't think it's spam.

   If you send email just to me that I deem to be of general interest, I will feel free to remail it anonymously to the email list along with my reply unless you explicitly tell me otherwise.

   The emails will be archived at the listserv website, and I will also archive them at http://www.cse.buffalo.edu/~rapaport/717cva/EMAIL/.

   For more information, read the Listserv Information webpage.

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).

For further information on my philosophy of grading, see my web document on "How I Grade"

Incompletes:

For more information on Incomplete policies, see the Graduate School web page, "Incomplete Grades".

ACADEMIC INTEGRITY:

(*) The idea and wording for this reading-commentary assignment are borrowed from the assignments for Stuart M. Shieber's course "Can Machines Think". [Back to text]