摘要
[目的/意义]针对目前用户兴趣变化的研究未考虑相邻时间段用户兴趣变化这一问题,在计算用户兴趣度的基础上,构建了用户兴趣变化模型,以准确把握用户的兴趣变化。[方法/过程]以小木虫论坛为例,基于时间衰减模型计算用户在各主题方向的兴趣度。在此基础上,构建基于用户兴趣历史相似度的兴趣变化模型,并采用LDA模型验证所构建模型的合理性。[结果/结论]该模型能够精确地刻画用户在相邻时间段对各主题方向的兴趣度变化情况。
[Purpose/significance]Against the problem of the current researches on user interest migration without considering the changes in user interest in adjacent time periods,the paper constructs a user interest migration model based on the calculation of user interest degree,so as to accurately grasp user interest changes. [Method/process] The paper takes the Xiaomuchong Forum for example,calculates the user interest degree in various theme directions based on time decay model,on this basis,constructs an interest migration model based on the historical similarity of user interest,and uses LDA model to verify the rationality of this model mentioned above. [Result/conclusion] The user interest migration model can accurately describe the changes of user interest in each topic direction in the adjacent time period.
作者
王险峰
Wang Xianfeng(Xinzhou Teachers University Library,Xinzhou Shanxi 034000)
出处
《情报探索》
2022年第9期111-116,共6页
Information Research
