LBSC 796/CMSC 828o
Information Retrieval Systems
Spring 2004
Course Description

Catalog Description

Goals

• Appreciate the capabilities and limitations of information retrieval systems.
• Understand the design and implementation of retrieval systems for text and other media.
• Evaluate the performance of an information retrieval system.
• Identify current research problems in information retrieval.

Approach

We will develop a general model for information retrieval systems in class, and will then use readings and class discussions to explore how that model can be used as a basis for understanding the design of information retrieval systems for a variety of applications. Homework assignments will be used to explore individual techniques in greater detail. Students will complete a term project in which they build a retrieval system, compare the performance of two systems, or conduct original research, all with the goal of developing a better understanding of implementation and evaluation issues.

Contact Information

Schedule

A syllabus that summarizes what we will cover each week can be found on the class Web site. There are no scheduled lab sessions. Students wishing to discuss accommodations for unusual circumstances should come see me before the third class session.

Course Materials

The required text for the course is Douglas W. Oard, Interactive Information Retrieval: An Interdisciplinary Perspective, which is available as a pre-publication draft on the password protected readings page.

Textbook and supplemental reading assignments for each week can be found on the syllabus. The required text, the assigned supplemental readings, and (except as noted) the following recommended texts will be on reserve in the Paul Wasserman Library on the second floor of the South Wing of the Hornbake Library building (note: this is not accessable through the main entrance to Hornbake Library).

• Ricardo Baeza-Yates and Berthier Rubiero-Neto, Modern Information Retrieval, Addison Wesley, 1999.
• Ian H. Witten, Alaitair Moffat, and Timmothy C. Bell, Managing Gigabytes, Morgan Kaufmann, Second Edition, 1999.
• David A. Grossman and Ophir Frieder, Information Retrieval: Algorithms and Heuristics, Kluwer Academic, 2004. Not yet available for the library, but preprint available on the password protected Web page.
• William B. Frakes and Ricardo Baeza-Yates, ed., Information Retrieval: Data Structures and Algorithms, Prentice-Hall, 1992.
• Tomek Strzalkowski, ed., Natural Language Information Retrieval, Kluwer, 1999.
• Christopher D. Manning and Heinrich Schuetze, Statistical Natural Language Processing, MIT Press, 2000.
• Karen Sparck-Jones and Peter Willet, ed., Readings in Information Retrieval, Morgan-Kaufmann, 1997.
• David C. Blair, Language and Representation in Information Retrieval, Elsevier Science, 1990.

Students are encouraged to use the course mailing list (lbsc796-0101-spr04@coursemail.umd.edu) to share information that would be of genral interest, to describe their project interests when seeking potential partners, or for any other purpose that seems reasonable. Mail sent to that address will reach the professor, TA, and all students. Students that were registered as of January 28, 2004 were automatically subscribed to the list using the email address that the university has on file (usually your WAM account). Please forward your email from there to where you actually read mail, or contact oard@umd.edu if you want your address on that list changed.

When possible, lectures will be videotaped and placed on reserve in the Paul Wasserman Library. These videotapes can be viewed there, but because there is only a single copy they cannot be checked out. RealVideo recordings of each class will also be made when possible and linked to from the syllabus. Students wishing to view a live RealVideo feed for specific sessions should make arrangements with the professor in advance.

In addition to your WAM account, CLIS students will also need a Glue account in order to complete the homework assignments and the term project. To get a Glue account, telnet to https://www.oit.umd.edu/cgi-bin/account/activation.cgi and follow the instructions that appear. CS students will be able to use their CS network accounts for class assignments, and therefore need not sign up for a glue account.

Grading

The following information regarding the ways in which CMSC 828o can count towards CS graduate program requirements is copied directly from the CS department's MS/PhD status of special topics courses page:

• CMSC 838O (Oard) Information Retrieval Systems (cross-listed with LBSC796)
  • Phd course in DB only if student uses CMSC 724 as the other course for DB Phd coursework requirement.
  • MS course in DB only if student successfully completes CMSC 724.
  • Final counts as MS comp in DB only if student successfully completes CMSC 724.

The homework and programming assignments are designed to provide an opportunity for students to explore specific topics in a structured way. Students may work together on the homework assignments, but all of the material that is turned in for grading must be produced individually. For example, LBSC 796 students may form study groups and work out homework solutions together on a chalkboard or by each working separately on different terminals and then sharing what they have learned, but it would not be permissible for one student to prepare an answer set and then for other students to copy those answers and submit it as their own work. CMSC 828o may work together to understand the design of an existing information retrieval system, to design the logic for a new module, or to debug a program, but each student must write any new code and perform any submitted runs individually.

A term project will be completed by the end of the semester. Students may work individually or in small groups. Additional details are provided on the term project Web page.