Improving Web Information Retrieval Using Shared Contexts

The effective utilisation of a user's context in improving the performance of web search engines is a subject of intense research interest. In particular, much attention has been directed to the enhancement of queries and the provision of more relevant information by taking user context into account. Progress in this field has been limited to date, however, due to ongoing challenges in capturing and representing contextual information. The authors describe here the development and evaluation of a web-based contextual information retrieval that addresses some of these challenges and makes progress in defining the information required to create contextual profiles. Their system collects and leverages implicit and explicit user data to modify queries with the aim of more accurately reflecting the user's interests.