基于查询\|概念的用户兴趣模型构建

吉林大学学报(信息科学版) ›› 2015, Vol. 33 ›› Issue (3): 310-314.

基于查询\|概念的用户兴趣模型构建

宋天勇, 赵辉, 郑山红, 王国春

长春工业大学计算机科学与工程学院, 长春 130012

收稿日期:2014-12-29 出版日期:2015-05-23 发布日期:2015-07-25
作者简介:宋天勇（1988—）, 男, 河南新乡人, 长春工业大学硕士研究生, 主要从事文本挖掘、智能计算研究, (Tel)86-18611832286(E-mail)1139026513@qq.com; 通讯作者：赵辉(1972—), 女, 吉林敦化人, 长春工业大学副教授, 博士, 主要从事智能计算、搜索引擎研究, （Tel）86-13074367546(E-mail)zhaohui@mail.ccut.edu.cn。
基金资助:
吉林省科技厅自然科学基金资助项目（20130101060JC)；吉林省教育厅“十二五”科学技术研究基金资助项目（2014132; 2014125）

Construction of User Interest Model Based on Queries and Concept

SONG Tianyong, ZHAO Hui, ZHENG Shanhong, WANG Guochun

College of Computer Science and Engineering, Changchun University of Technology, Changchun 130012, China

Received:2014-12-29 Online:2015-05-23 Published:2015-07-25

摘要/Abstract

摘要：

针对查询\|概念二分图因概念抓取和查询词权重设计不足而导致构建的用户兴趣模型不合理的问题, 提出一种基于查询\|概念二分图的用户兴趣建模算法。通过tf×idf公式抓取概念, 并利用用户对查询词的浏览时间计算查询词的权重, 确保改进后的查询\|概念二分图能更准确地表示用户的查询意图。实验结果表明, 该算法构建的用户兴趣更为合理。

关键词: 查询-概念二分图, 浏览时间, 用户兴趣模型

Abstract:

Bipartite graph of queries and concept is insufficient in concept grabbing and calculating weight of the query words, which cause the user interest model construction not reasonable. The improved bipartite graph of queries and concept will represent the users query intent more accurate by using tf×idf grasping concepts and calculating the weight of query query words through user's browsing time. A user interest modeling algorithm based on improved bipartite graph of queries and concept was proposed. And experiments verified that user interest model constructed by this algorithm is more reasonable.

Key words: bipartite graph of queries and concept, browsing time, user interest model （UIM）

中图分类号:

TP301

宋天勇, 赵辉, 郑山红, 王国春. 基于查询\|概念的用户兴趣模型构建[J]. 吉林大学学报(信息科学版), 2015, 33(3): 310-314.

SONG Tianyong, ZHAO Hui, ZHENG Shanhong, WANG Guochun. Construction of User Interest Model Based on Queries and Concept[J]. Journal of Jilin University(Information Science Ed, 2015, 33(3): 310-314.