| 邹婧琳,范炜.基于网络读书社区阅读行为的书目推荐模型研究[J].数字图书馆论坛,2014,(4):49~55 |
| 基于网络读书社区阅读行为的书目推荐模型研究 |
| Bibliography Recommendation Model Derived from Reading Behaviors of Online Social Book Community |
| |
| DOI： |
| 中文关键词: 网络读书社区;阅读行为;书目推荐 |
| 英文关键词: Online social book community;Reading behaviors;Bibliography recommendation |
| 基金项目:本研究得到四川大学大学生创新创业训练计划项目“大学生书目导读与互动推荐系统”（编号20131241）资助。 |
|
| 摘要点击次数: 1810 |
| 全文下载次数: 1299 |
| 中文摘要: |
| 书目推荐旨在建立图书与读者之间的双向连接。新兴网络读书社区中用户的参与互动过程积累了大量用户行为数据，为书目推荐提供了丰富的数据依据。文章针对网络读书社区中的主要用户阅读行为，综合考虑用户评分、阅读状态、标签、时间信息对用户阅读偏好的体现，提出用户阅读偏好程度和偏好方向相结合的用户偏好模型，形成了个性化书目推荐。采集豆瓣读书数据进行模型检验，结果表明，该模型有效地提高书目推荐的准确性和多样性，对网络读书社区实施个性化书目推荐具有参考价值。 |
| 英文摘要: |
| Bibliography recommendation aims to build a two-direction connection between books and readers. Due to users' participation in the online social reading community, huge data is newly created every day, which provides many references for bibliography recommendation. This paper analyzes users' primary reading behaviors and studies how users' rating, reading state, tag and action time present their reading preferences. Then it proposes a bibliography recommendation model, combining with users' reading preference intentions and levels. Final y, it uses the dataset from douban.com to evaluate the recommendation model. The experimental results show that the model achieves bet er recommendation in terms of accuracy and diversity, and the model could provide an action reference for online social reading community. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
