Research on Search System Based on Agent and Personal Knowledge Ontology

There is a lack of semantic comprehension in the traditional information search models. Now, Ontology provides a semantic-based knowledge representation and sharing which is a formal, explicit specification of a shared conceptual model. This paper introduces personal knowledge ontology, comprehensively utilizes various characteristics of an intelligent Agent such as initiative, collaboration, mobility and so on, and adopts the basic principle of collaborative filtering, and then puts forward the search system based on Agent and personal knowledge ontology, which greatly improves the search efficiency. Meanwhile, the paper presents an algorithm which is used to revise the concept of personal knowledge ontology through constant study of the feedback information from users, and also explores the application of multi-Agent collaborative information filtering in search system.