元搜索引擎的个性化调度算法  

Improved Scheduling Algorithm Based on User Profile for Meta Search Engine

作　　者：毛佳[1] 康慕宁[1] 

机构地区：[1]西北工业大学计算机学院,西安710072

出　　处：《微处理机》2008年第6期125-127,共3页Microprocessors

基　　金：国家自然基金项目(60573096)

摘　　要：元搜索引擎的调度算法是研究如何从庞杂的独立搜索引擎中选择出与查询字串相关度最高、与用户的查询需求最贴近的合适数量的独立搜索引擎。现在,在原有的元搜索引擎调度算法基础上,提出了一种个性化调度算法。该算法根据用户兴趣类对所有独立搜索引擎进行文档分类,然后根据用户查询串所属的兴趣分类,计算出查询串与该分类下文档的相关度这一调度算法的主要影响因素,再结合成员搜索引擎的平均响应时间性能评价,返回结果数量,以及以用户反馈为基础的用户兴趣度经验,计算出独立搜索引擎的排序,从而实现个性化的调度。Scheduling algorithm of meta - search engine is to help users choose the proper ones from lots of search engines, which have higher relativity with query string, and meet user＇ s searching requirement better. A kind of personalized algorithm is proposed, which improved customary scheduling algorithm. It classifies the files in the database of every search engine according to user profile. Then, compute the similarity between the query string and classified files, which share the same interest classification with query string. Also considering average respond time of every search engine, the number of results, user＇ s experience based on users＇ feedback. We can get the ranking of every search engine to realize personalized scheduling strategy.

关 键 词：元搜索引擎 成员搜索引擎(独立搜索引擎) 调度策略 相关度 用户兴趣 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]