Building of searching behavior analysis models on multi-agent intelligent agent technology

The traditional keywords or server log based searching behavior analysis is usually simple which makes it difficult to meet the individual requirements of the user information search. This thesis presents a modeling method of searching behavior analysis based on multi-agent technology, in which the searching behavior analysis is divided into four kinds, including user search keywords, server logs information, user downloading and saving information, the visit time on the web pages. Based on which the user search behavior and the model of user-search behavior analysis are automatically built to well express the users´ interests and realize the user individual services to improve the search engine efficiency.