Personal Adaptive Web agent: a tool for information filtering

This paper highlights the problem of information overload and recall on the WWW. It briefly mentions four categories of solutions which are: Search Engines, Personal Homepages, Information Management Tools, and Information Retrieval Agents. The paper also presents our own framework of a Personal Adaptive Web (PAW) agent and its current results. Our agent uses a real-time database, automatic text analysis, a competitive learning network, and fuzzy logic to learn the different categories of Web documents that the user is interested in, then finds and suggests new similar documents.