学术虚拟社区核心用户兴趣迁移模型构建研究  

Research on the Construction of Core User Interest Transfer Model in Academic Virtual Community

作　　者：杨瑞仙 黄书瑞 王彰奇[3] Yang Ruixian;Huang Shurui;Wang Zhangqi(School of Information Management,Zhengzhou University,Zhengzhou 450001,China;Zhengzhou Data Science Research Center,Zhengzhou 450001,China;Puyang Institute of Science and Technology Information,Puyang 457000,China)

机构地区：[1]郑州大学信息管理学院,河南郑州450001 [2]郑州市数据科学研究中心,河南郑州450001 [3]濮阳市科学技术情报研究所,河南濮阳457000

出　　处：《现代情报》2021年第2期10-18,共9页Journal of Modern Information

基　　金：国家社会科学基金项目“学术虚拟社区知识交流效率测度研究”(项目编号:17CTQ030);河南省高等学校青年骨干教师培养计划“学术虚拟社区知识交流机制的系统动力学仿真研究”(项目编号:2020GGJS012);郑州大学人文社会科学优秀青年科研团队培育计划“用户参与的网络知识服务研究”(项目编号:2020-QNTD-09)。

摘　　要：[目的/意义]针对目前学术虚拟社区存在对核心用户兴趣变化关注不足的问题,本文基于艾宾浩斯遗忘曲线构建用户兴趣迁移模型,以准确描述核心用户的兴趣迁移,提升对核心用户个性化推荐结果的准确率。[方法/过程]以"小木虫论坛"为研究对象,构建用户影响力评估指标和问答网络,分别利用熵权法和PageRank算法识别核心用户。在此基础上,基于艾宾浩斯遗忘曲线构建核心用户兴趣迁移模型,并对此模型的适用性进行验证。[结果/结论]本文所构建的核心用户兴趣迁移模型能更好地反映核心用户的兴趣迁移。[Objective/Meaning] Aiming at the current academic virtual community’s insufficient attention to changes in core user interest,this paper builds a user interest transfer model based on the Ebbinghaus forgetting curve to accurately describe the core user’s interest transfer and improve the accuracy of personalized recommendations for core users. [Method/Process] The paper took the "emuch BBS" as the research object,constructed user influence evaluation indicators and question-and-answer network,and then respectively used entropy method and PageRank algorithm to identify core users. On this basis,the core user interest transfer model was constructed based on the Ebbinghaus Forgetting Curve,and the applicability of this model was verified. [Results/Conclusions] The core user interest transfer model constructed in this paper could better reflect the core user’s interest transfer.

关 键 词：学术虚拟社区 用户兴趣迁移模型 小木虫 熵权法 PAGERANK算法 核心用户识别 艾宾浩斯遗忘曲线 协同过滤算法 

分 类 号：G203[文化科学—传播学]