摘要
Similarity matching and information presentation are two key factors in information retrieval.In this paper,a saliency-based matching algorithm is proposed for user-oriented search based on the psychological studies on human perception,and major emphasis on the saliently similar aspect of objects to be compared is placed and thus the search result is more agreeable for users.After relevant results are obtained,the cluster-based browsing algorithm is adopted for search result presentation based on social network analysis.By organizing the results in clustered lists,the user can have a general understanding of the whole collection by viewing only a small part of results and locate those of major interest rapidly.Experimental results demonstrate the advantages of the proposed algorithm over the traditional work.
Similarity matching and this paper, a saliency-based information presentation are two matching algorithm is proposed key factors in information retrieval. In for user-oriented search based on the psychological studies on human perception, and major emphasis on the saliently similar aspect of objects to be compared is placed and thus the search result is more agreeable for users. After relevant results are obtained, the cluster-based browsing algorithm is adopted for search result presentation based on social network analysis. By organizing the results in clustered lists, the user can have a general understanding of the whole collection by viewing only a small part of results and locate those of major interest rapidly. Experimental results demonstrate the advantages of the proposed algorithm over the traditional work.
基金
Supported by the Fund for Basic Research of National Non-Profit Research Institutes(No.XK2012-2,ZD2012-7-2)
the Fund for Preresearch Project of ISTIC(No.YY201208)
关键词
匹配算法
搜索结果
集群
浏览
面向用户
信息检索
网络分析
相似度
information retrieval, saliency-based matching, human perception, cluster-based browsing, social network