Grouper home site

Overview

Clustering is an effective way of organizing documents into collections for ease of browsing. Recently with the growth of WWW, clustering has become a paradigm for organizing search results. Online systems face many new challenges, including the need for fast response time, generating high quality clusters with simple descriptions for novice users, and working with document distributions that violates many traditional assumptions. How do different clustering algorithms trade off quality of clusters and speed? What modifications are necessary to adapt traditional clustering algorithm to the WWW? How do these system scale to larger document collection? How do these systems evaluate the quality of the cluster they generate? How are the clusters generated in each case, and are there any processing after cluster generation to improve on the cluster quality?

Publications

• Zamir. Clustering Web Documents: A Phrase-Based Method for Grouping Search Engine Results. Ph.D. Thesis. 1999.
• Zamir and Etzioni. Grouper: A Dynamic Clustering Interface to Web Search Results. In WWW8. 1999.
• Landau, Feldman, Aumann, Fresko, Lindell, Sipshtat, and Zamir. TextVis: An Integrated Visual Environment for Text Mining. In PKDD '98. 1998.
• Zamir and Etzioni. Web Document Clustering: A Feasability Demonstration. In SIGIR'98. 1998
• Zamir, Etzioni, and Karp. Fast and Intuitive Clustering of Web Documents. In KDD '97. 1997.