CAJ | 학술논문

搜索引擎在多成员搜索引擎搜索结果的整合过程中，搜索结果的排序在很大程度上决定着元搜索引擎的服务质量。为了实现搜索结果的有效整合，目前技术主要结合查询请求、文档内容、初始排序或（和）赋予搜索成员搜索引擎权重等因素。其中采用赋予搜索引擎权重时，往往根据用户和技术人员经验，主观地进行赋值，不能体现真实的用户搜索偏好。为此，提出了通过挖掘用户搜索及遍历情况，动态地赋予各成员搜索引擎权重的方法。通过用户遍历及点击下载情况，得到了用户搜索遍历与返回结果的匹配度，论证了该方法的可行性和有效性。
수색인경재다성원수색인경수색결과적정합과정중，수색결과적배서재흔대정도상결정착원수색인경적복무질량。위료실현수색결과적유효정합，목전기술주요결합사순청구、문당내용、초시배서혹（화）부여수색성원수색인경권중등인소。기중채용부여수색인경권중시，왕왕근거용호화기술인원경험，주관지진행부치，불능체현진실적용호수색편호。위차，제출료통과알굴용호수색급편력정황，동태지부여각성원수색인경권중적방법。통과용호편력급점격하재정황，득도료용호수색편력여반회결과적필배도，론증료해방법적가행성화유효성。
In the integration process of search results returned from multi member search engines, the quality of a service meta-search engine is determined by the ranking of search results to a great extent. The current technologies on achieving effective integration of search results mainly rely on combining search queries, document content, the initial sorting infor-mation or(and)endowing the member search engines with weights, and others. The method of endowing engines with weights is objective, often depends on the subjective experience of users, and can’t embody users’search preferences. So, this paper proposes a method to dynamically endow the member search engines with weights based on mining user’s search-ing and navigation habits. It analyzes users’log data on clicking and downloading search results, and gets the matching degrees between search results and users’navigation. The statistic results demonstrate that the methodology is feasible and effective to the ranking of multi engine search results.