Reducing the Cost of Knowledge

Abstract

This thesis will examine the process of sensemaking and the costs involved, and present and evaluate a prototype document-oriented web information workspace that aims to reduce these costs. The user interface of this prototype system enables various forms of organization with the intent of increasing the efficiency of locating documents in the workspace. Furthermore, it will propose a system that aggregates the organizational structure of multiple users to form a collective mental model of the web. In traditional IR, user information need is defined only by the query. By providing the query's context, more information about the user information need is provided. The context is used to search the collective mental model for relevant documents, and used to modify the ranking obtained from a more traditional keyword-based search.