图书馆杂志

图书馆杂志 ›› 2013, Vol. 32 ›› Issue (8): 36-41.

• 工作研究 • 上一篇    下一篇

基于检索行为的非个性化图书推荐

  黎邦群   

  • 出版日期:2013-08-15 发布日期:2013-08-25
  • 通讯作者: 黎邦群 惠州学院图书馆, 副研究馆员。 广东 516000

Non-personalized Book Recommendation Based on Search Behavior

  Li Bangqun   

  • Online:2013-08-15 Published:2013-08-25

摘要:

针对当前OPAC缺少大众化的图书推荐服务而造成图书推荐精准度低下、资源利用率不高的
现状,进行非个性化图书推荐的应用现状与需求分析、技术设计与功能实现的研究。通过设计方
案、分析思路、阐述原理及构建数据表并对用户检索及结果选择行为特征及其隐含的关联规则进行
知识分析,在此基础上存储、查询行为数据,挖掘出图书的相关度或相似性,实现非个性化图书推
荐功能,并通过权重、显示、排序等优化策略来提升用户使用体验,以期提高图书的可发现性,提
升图书馆的服务质量。

关键词: 图书推荐 , 非个性化 , OPAC, 用户行为 , 检索行为 , 相关度 , 相似性

Abstract:

To solve the problems of poor precision of book recommendation and low resource utilization rate
when OPAC lacks popular books recommendation, the paper studies the status quo of non-personalized book
recommendation, technical design and potential measures for improvement. It expounds the principle, builds
the data tables, and analyzes the correlation between users’ search behavior and final choices. Based on these
statistics, it suggests the system should save and search behavioral data so as to work out the correlation and
similarity between books. Through optimization strategy, it could thus improve users’ experience, enhance the
rate of utilization of books in library and advance the library service.

Key words: Books recommendation, Non-personalized, OPAC, Search behavior, Correlation, Similarity