THE IMPLEMENTATION OF AN INTELLIGENT SEARCH AGENT CHEN Hongying

摘要

This paper introduces a full-scale intelligent search engine, which has used several machine-learning techniques. Our engine has several periods: search period, filter period and interest-learning period. In search period we use concept search method and multi-thread method; in filter period, we use improved Vector Space Model method; in interest-learning period we use heuristic learning method, decision tree learning method and Initiative service method.