SWAMI: Searching the Web Using Agents with Mobility and Intelligence

The rapid growth of the World Wide Web has complicated the process of web browsing by providing an overwhelming wealth of choices for the end user. To alleviate this burden, intelligent tools can do much of the drudge-work. This paper describes the SWAMI system. It combines multiple aspects of adaptive web technologies into a framework for an intelligent web browsing system. It uses a multi-agent system to represent the interests of the user dynamically and takes advantage of the active nature of agents to provide a platform for parallel look-ahead evaluation, page searching, and cooperative link recommendation swapping. The collection of agents reflects the user’s interests by self-organizing into a hierarchicy according to the evidence of apparent interest demonstrated by the user. Example results of the functioning prototype are presented, demonstrating its ability to infer and react to a user’s interests.

• Type: Conference
• Year: 2005
• Where: AI
• Authors: Mark Kilfoil, Ali A. Ghorbani
• Keywords: AI 2005, Artificial Intelligence, User’s Interests, Web Browsing, Web Browsing System