Abstract: 【Purpose/significance】This paper aims to identify the hot topics of information privacy research and summarize the evolu⁃tion trajectories of research topics【. Method/process】Considering the research limitations of topic identification and evolution,this pa⁃per proposed a comprehensive research method of topic evolution from the perspective of temporal correlation and structural represen⁃tation.First,the LDA (Latent Dirichlet Allocation) model is applied to identify the research topics of scientific publications in multiple temporal intervals.Then,the topic condensed subgroups are depicted based on the co-word analysis to get better semantic performance.Furthermore,the evolutionary paths of different topics under adjacent intervals are explored from the dimension of temporal correlation.Moreover,several scientometric indicators,such as novelty,centrality and influence,are defined to investigate the changes of emergent
topics and hot topics in information privacy field from the perspective of structural representation【. Result/conclusion】The empirical results show that the proposed method can identify the research topics effectively, and analyze the topic evolution trends from both macro and micro dimensions.This study can expand the theoretic perspective of topic evolution research,and detect the emergence and hotspots of information privacy research more efficiently【. Innovation/limitation】An integrated research framework of LDA topic model and co-word analysis is proposed to identify the research topic more precisely,and the topic evolution path is analyzed from the per⁃
spective of both temporal correlation and structural representation.In future study,multiple data sources and multi-group terms will be introduced to improve the research method.