摘要
元搜索引擎并行地向各个成员搜索引擎发出请求,合并及处理所有成员引擎的返回结果。相对于传统搜索引擎,元搜索引擎具有更好的查全率但在结果相关度排序及查准率方面仍需要改善。就相关度排序及查准率方面的问题元搜索成员引擎对于各个不同主题具有不同的检索质量并就此提出一种基于主题偏好的排序方法。利用Beeferman聚类方法对检索主题划分,通过Borda排序算法对元搜索引擎获得条目进行基于主题的分类排序,以此来提高元搜索查询质量和改善用户体验。
Meta-search engine launches query simultaneously to its member search engines and shows a combined and ranked results list. Compared with the traditional search engine,meta-search engine has a bet^r recall rate. However with the large amount of return items from its member engines,the precision rate and MMR still need to be perfected. Each member engine performs differently in the searching tasks with different topics in view of precision rate and MMR. In this paper, present a topic preference based ranking algorithm. Using Beeferman clustering method divides the search topic, with Borda ranking algorithm classify and rank the entries obtained by meta-search engine based on topic,improving the meta-search query quality and enhancing user experience.
出处
《计算机技术与发展》
2013年第2期41-43,48,共4页
Computer Technology and Development
基金
国家自然科学基金资助项目(60673186)
江苏省高校"青蓝工程"中青年学术带头人培养对象资助项目
关键词
元搜索引擎
主题偏好
排序算法
聚类
meta-search engine
topic preference
ranking algorithm
clustering