基于关键词设置的P2P信息检索

P2P INFORMATION RETRIEVAL BASED ON KEYWORDS SETTING

在分析以往结构化P2P基于单关键词分割的信息检索基础上,结合利用用户频繁出现的请求关键词的特点,提出结构化P2P中的一个基于请求关键词设置的索引框架KS-PIR。采用通过关键词联合的新特征项设置来扩充索引特征,并删节与特征项相关的索引记录表至恒定长度的方法,在检索质量提高的同时,又保证了可接受的带宽消耗,系统具有很好的扩展性。

Previous structurized P2P information retrieval based on single-keyword segmentation is analysed. On that basis, by combining and utilising the character of keywords queries frequently occurred from users,an indexing framework based on the setting of kcywords queries （KS-PIR） in structurized P2P is presented, It adopts the approaches of extending index features by the setting of new feature items with keywords combination,and of truncating the index log associated with features to a constant size. The new system improves the quality of the retrieval,at the same time,it ensures the acceptable bandwidth consumption,and has good scalability.