计算机科学 ›› 2012, Vol. 39 ›› Issue (3): 157-159.

• 数据库与数据挖掘 • 上一篇    下一篇

基于信息量与信息嫡的元搜索引擎排序算法研究

赖相旭,韩立新,曾晓勤,王敏,吴胜利   

  1. (河海大学计算机与信息学院 南京210024);(南京大学计算机软件新技术国家重点实验室 南京210093);(阿尔斯特大学计算机及信息科学学院 英国贝尔法斯特)
  • 出版日期:2018-11-16 发布日期:2018-11-16

Research of Ranking Algorithm Based on Information Quantity and Entropy in Meta Search Engine

LAI Xiang-xu,HAN Li-xin,ZENG Xiao-qin,WANG Min,WU Sheng-li   

  • Online:2018-11-16 Published:2018-11-16

摘要: 元搜索引擎集合了多个成员搜索引擎的结果,将结果进行一定的处理后再将处理后的结果返回给用户。其中对结果的重新排序直接影响到元搜索引擎的性能。基于通信领域上的信息量与信息嫡提出一种计算结果相关度的算法—信息关联度IRI)算法,再将算法进行特定的修正,并提出一种合并算法CombMul,将以上算法应用到元搜索引擎中,最终用MRR查准率来评价此方法。得到的MRR查准率数据表明,与广泛应用的I3orda排序算法相比,IRD算法结果更为理想。

关键词: 元搜索引擎,排序算法,信息关联度,IRIJ,信息量,信息嫡,CombMul

Abstract: The meta search engine collects results from many search engines, using a certain way to treat the results and then returning back to the users. Reranking the results will directly affect the performance of meta search engine. This paper was based on information quantity and entropy which arc used in communication field then presented a calculation algorithm-information related degree(IRD),after a particular amendment to the IRl)algorithm, this paper also pro- posed a merging algorithm combMul. The above algorithms were applied to the meta search engine, and MRR precision was used to evaluate the algorithm .The MRR precision data show that IRD algorithm is even better compared with a widely used sorting algorithm-Borda.

Key words: Meta search mngine, Ranking algorithm, Information related degree, IRD,Information quantity, Entropy,ComhMul

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!