摘要
提出一种基于用户行为聚类的搜索引擎算法。该算法从用户行为日志中挖掘用户意图,并根据用户的反馈信息定位用户意图信息,提升了查询的准确率,有效地解决了传统的全文检索式搜索引擎查询具有二义性词时的不足,并通过实践验证了算法的可行性。
This paper proposes an algorithm and an instance of the search engine based on clustering users’ activities. The algorithm mines users’ concepts from activities log and uses users’ feedback to identify the concepts. Both of them improve the accuracy of the engine. It resolves the ambiguity problem of full-text retrieval search engine. At last,the paper gives a case to implement the algorithm.
出处
《计算机系统应用》
2010年第4期94-97,共4页
Computer Systems & Applications
