Binary Information Press
The information on Internet expands so fast, so how to provide effect and sufficient information to meet the customers' satisfactory has become a frontier in Information Retrieval. Personalized information retrieval provides search results based on the unique customer's characteristic. The key difficult in the long-term personalized information retrieval is how to depict the user's characteristic. This paper analyzes the user behavior characteristics on Internet and proves the feasibility for the user profile model built in search engine. Take the effect of Ontology adopted in the semantic search into consideration, the experiment combines the Ontology technology and the user long-term historical behavior data to produce the user profile model.