企业搜索引擎个性化表示与结果排序算法研究

发布时间：2018-03-25 10:39

本文选题：本地文档集　切入点：个性化　出处：《计算机研究与发展》2014年01期

【摘要】：针对企业搜索引擎提出一种基于本地文档库的个性化表示与结果排序算法,以帮助用户找到真正感兴趣的结果.首先,采用聚类分析对用户浏览的历史文档聚类;其次,采用模糊推理技术对所形成的分类进行分析,发现用户对各分类的喜好程度;再次,按用户对各分类喜好程度的不同,为各分类分配抽样文档数;最后,采用多种抽样技术,从各分类中抽取典型文档.来自不同分类的典型文档构成了表示用户个性的本地文档库.结果排序算法通过计算通用企业搜索引擎的搜索结果与本地文档库中各文档的相似性,对结果集重新排序,从而体现出用户个性.实验结果表明,与传统的基于关键词的个性化表示与结果排序算法相比,基于本地文档集的个性化表示与结果排序算法可以给出更能反映用户个性的查询结果,且可以对用户偏好的变化作出更迅速的反映。
[Abstract]:A personalized representation and result sorting algorithm based on local document library is proposed for enterprise search engines to help users find the results they are really interested in. Firstly, clustering analysis is used to cluster the historical documents viewed by users; secondly, Fuzzy reasoning technology is used to analyze the resulting classification, and the user's preference for each classification is found. Thirdly, according to the user's different preferences for each classification, the number of sampling documents is allocated for each classification. Finally, a variety of sampling techniques are used. Typical documents are extracted from each classification. Typical documents from different categories constitute a local document library that represents the personality of the user. The result sorting algorithm calculates the similarity between the search results of the general enterprise search engine and the documents in the local document library. The result set is reordered to reflect the personality of the user. The experimental results show that, compared with the traditional keyword based personalized representation and result sorting algorithm, The personalized representation and result sorting algorithm based on the local document set can give the query results which can reflect the user's personality more quickly, and can also reflect the change of the user's preference more quickly.
【作者单位】：厦门大学软件学院;
【基金】：国家自然科学基金项目(61100032) 中央高校基本科研业务费专项基金项目(2010121072,2010121070)
【分类号】：TP391.3

【共引文献】