The objective of the retrieval project is to provide you with an opportunity to study in depth aspects of on-going research in the information retrieval area. In fact, some of the topics below are dissertations at other schools.

This project is to be accomplished in a group. Ideally, you should form groups of three each. If you have a compelling reason why a group with less than three is ideal in your case, please let me know immediately. In any case, you should have the group as well as the topic cleared with me before proceeding. It is important that you start with the project as soon as possible, even if we have not directly covered the material in the class (due to a quirk in scheduling, retrieval is covered rather late in the semester).

You should study the topics by reading the basic paper/dissertation, and any work that you find in the related areas. I shall be posting the guidelines for the group oral presentations as well as written report on the project during the semester.

All of the following are available over the web, if you are resourceful in finding them.

• Content Routing: A Scalable Architecture for Network-Based Information Discovery, by Mark A. Sheldon (Ph. D Dissertation, EECS, MIT, 1995).
• Identifying and Merging Related Bibliographic Records, by Jeremy A. Hylton (M.E. Dissertation, EECS, MIT, 1996).
• Determining Similarity and Inferring Relations in a Lexical Knowledge Base, by Stephen D. Richardson (Ph. D Dissertation, CS, CUNY, 1997). Microsoft
• Inference Networks for Information Retrieval, by Howard R. Turtle (Ph. D Dissertation, CS, U Mass, 1991). U Mass
• Information Extraction as a Basis for Portable Text Classification Systems, by E. Riloff (Ph. D Dissertation, CS, U Mass, 1995). U MASS
• WORDNET: An On-Line Lexical Database, by George A. Miller (Princeton, 1993). Princeton
• Scatter/Gather: A Cluster-based Approach to Browsing Large Document Collections, by Douglas A. Cutting, David R. Karger, Jan O. Pedersen, and John W. Tukey (XEROX/PARC, 1992). XEROX/PARC