The Adaptive Webbrowser: Motivation

Name	The Adaptive Webbrowser: Motivation
Description	This section is based on Syskill & Webert, by M. Pazzani and D. Billsus. Here we focus on the problem of assisting a person to find information that satisfies long-term, recurring goals (such as finding information on machine applications in medicine) rather than short-term goals (such as finding a paper by a particular author). The "interestingness" of a webpage is defined as the relevance of the page with respect to the user's long term information goals. Feedback on the interestingness of a set of previously visited sites can be used to learn a profile that would predict the interestingness of unseen sites. In this section, it will be shown that revising profiles results in more accurate classifications, particulary with small training sets. For a general overview of the idea, see Figure 4: Figure 4. The user classifies visited sites into "interesting" and "not interesting". This website collection is used to create a user profile, which is used by the adaptive webbrowser to annotate the links on a new website. The results are shown to the user.
Case Study	The Adaptive Webbrowser Case