基于语义主题图谱的学术APP用户信息需求发现研究

Research on Information Requirement Discovery of Academic APP Users Based on Semantic Topic Map

[目的/意义]提出了一种基于语义主题图谱发现学术APP用户信息需求内容的思路方法,充分发挥语义主题图谱的深层关联优势,为运营者精准感知用户信息需求提供指导。[方法/过程]首先从提问数据中筛选出需求文本,然后通过TF-IDF、LDA主题模型以及词性限制的方式抽取需求主题词,最后基于Glove模型从全局语义角度挖掘主题词的关联词并生成语义主题图谱,利用语义主题图谱的语义共现、语义扩展、语义集成作用辅助运营者发现用户信息需求。以小木虫APP为例进行实证分析。[结果/结论]实验结果表明,利用该方法能构建出深层次、强关联的语义主题图谱,运营者应用语义主题图谱可以发现用户核心需求焦点、具体需求内容和需求主题类别,对提升需求分析及决策效率有重要意义。

[Purpose/significance]This paper proposes an idea and method to discover the content of academic app users’information demand based on semantic topic map,which gives full play to the deep correlation advantages of semantic topic map,and provides guidance for operators to accurately perceiveusers’information demand.[Method/process]Firstly,the requirement text is selected from the query data,and then the requirement subject words are extracted by TF-IDF,LDA topic model and part of speech restriction.Finally,based on the glove model,the associated words of the subject words are mined from the perspective of global semantics,and the semantic topic map is generated.The semantic co-occurrence,semantic extension and semantic integration of the semantic topic map are used to assist the operators to find the user information demand Take the Xiaomuchong application as an example for empirical analysis.[Result/conclusion]The experimental results show that the deep level and strong correlation semantic topic map can be constructed by using the ideas and methods in this paper,and the operators can find the user’s core demand focus,specific demand content and demand topic category by using the semantic topic map,which is of great significance to improve the efficiency of demand analysis and decision-making.