Research on Results Merging in Meta Search Engine
Hongmei Li, Zhenguo Ding · Beijing Youdian Xueyuan xuebao · 2008
A result merging method based on text/rank analysis and group decision making activity is proposed.Utilizing text-based information obtained from search results,a definition of query-match grade is presented,and an approach on text normalization for meta search is described.The relevant scores of relevant documents are normalized by incorporating text analysis measure with existing rank- based method.When estimating scores of non-relevant documents,different assumptions are dis- cussed,and an improved shadow document method is proposed as well.So a merging method based on group decision making activity is adopted to sort the search results.Four different search engines are tested in the practical web environment.The experimental results show that this method is more effec- tive than other three merging methods:Round-robin,CombSum and CombMNg.