Abstract

To facilitate users' ability to make sense of large collections of hypertext we present two new techniques for inducing clusters of related documents on the World Wide Web. Users' ability to find relevant information might also be enhanced by finding lawful properties of document behavior and use. We present models and analyses of document use and change for the World Wide Web. Keywords Clustering, categorization, co-citation analysis, World Wide Web, hypertext, survival analysis, usage models INTRODUCTION The ever-increasing universe of electronic information competes for the effectively fixed and limited attention of people. Both consumers and producers of information want to understand what kinds of information are out there, how desirable it is, and how its content and use change through time. Our work aims to discover empirical regularities of hypertext content, use, and structure, and ways of exploiting these regularities to provide new ways of helping people to find and make...

Citations