图书情报工作 ›› 2023, Vol. 67 ›› Issue (12): 39-49.DOI: 10.13266/j.issn.0252-3116.2023.12.004

• 工作研究 • 上一篇    下一篇

基于认知升级理论的高校图书馆智慧推荐服务研究

杜丰瑞, 岳铁骐, 张彤阳, 徐健   

  1. 中山大学信息管理学院 广州 510006
  • 收稿日期:2022-11-30 修回日期:2023-02-06 出版日期:2023-06-20 发布日期:2023-07-06
  • 通讯作者: 徐健,副院长,教授,博士,博士生导师,通信作者,E-mail:issxj@mail.sysu.edu.cn。
  • 作者简介:杜丰瑞,硕士研究生;岳铁骐,硕士研究生;张彤阳,博士研究生。
  • 基金资助:
    本文系国家社会科学基金面上项目“基于知识发现的科研人员老龄化现象、演化规律及成因分析”(项目编号: 18BTQ076)研究成果之一。

Research on Intelligent Recommendation Service of University Libraries Based on Cognitive Upgrading Theory

Du Fengrui, Yue Tieqi, Zhang Tongyang, Xu Jian   

  1. School of Information Management, Sun Yat-sen University, Guangzhou 510006
  • Received:2022-11-30 Revised:2023-02-06 Online:2023-06-20 Published:2023-07-06

摘要: [目的/意义] 将认知升级理论融入图书馆智慧推荐服务中,以实现知以藏往、见贤思齐的智慧化推荐服务。 [方法/过程] 首先,从兴趣热度、内容质量评价和专指度 3 个指标入手,构建图书馆智慧推荐系统的指标体系;其次,基于认知升级理论,将用户分为“前辈”和“后辈”,通过改进协同过滤推荐算法计算用户相似度,将“前辈”的成功学习路径推荐给相似的后辈;最后,利用精准率、召回率、 AUC、 MMR、 F1 值等指标对离线实验和在线实验结果进行检验。 [结果/结论] 实验结果表明,改进后的智慧推荐算法相比传统协同过滤算法的实现效果有明显提高;对比离线实验和在线实验结果发现,在线实验的推荐效果显著提升,意味着若将基于认知升级理论的智慧推荐服务加以推广,将会对高校学生的专业素质培养和认知层次升级产生积极影响。

关键词: 认知升级理论, 智慧推荐, 高校图书馆, 智慧服务

Abstract: [Purpose/Significance] This study proposes to integrate the theory of cognitive upgrading into the intelligent recommendation service of the library, so as to realize the intelligent recommendation service of knowing to accumulate learning experience from others. [Method/Process] Firstly, this study constructed an indicator system of the intelligent recommendation system of the library from three dimensions: interest heat, content quality evaluation and specificity; Secondly, users were divided into ‘predecessors’ and ‘descendants’ based on cognitive upgrading theory. The user similarity was calculated by improving the collaborative filtering recommendation algorithm, and the successful learning path of ‘predecessors’ was recommended to similar ‘descendants’; Finally, the offline experiment and online experiment results were tested by using accuracy, recall, AUC, MMR, F1 value and other indicators. [Result/Conclusion] The experimental results show that the implementation effect of the improved intelligent recommendation algorithm is significantly improved compared with the traditional collaborative filtering algorithm. Especially, the comparison between offline experiment and online experiment shows that the online experiment has a remarkable recommendation effect. Our results indicate that if the intelligent recommendation algorithm based on the cognitive upgrading theory is promoted, it will have a positive impact on the professional quality training and cognitive level upgrading of students in the college scene.

Key words: cognitive upgrading theory, intelligent recommendation, university library, intelligent service

中图分类号: