针对目前大部分搜索引擎不能精确识别使用不同查询词而期望获得不同查询结果的情况,该文提出一种基于用户行为模型的搜索引擎的思想,论述其原型系统SEB实现过程和关键技术,其中的行为模型结合人类行为学相关理论对用户访问行为进行分类和表示,对搜索结果进行了处理,实现了个性化搜索。实验表明,经SEB原型系统处理的搜索结果更加符合用户需求。
When searching, different users may have different search purposes even if they use the same keywords. However, currently most search engines can not solve this problem. In this paper, the idea of creating search engine based on users' behavior model is proposed, and the implementing process and key techniques of the prototype SEB(Search Engine on Behavior) based on the idea are expatiated. The behavior model imports the theory of behavior science for classifying and expressing the users' accessing behaviors, and presses the returned results according to the classification and expression, and achieves the purpose of personalizadon. The experiments show that returned results processed by SEB can meet the uses' requirements better.