时序关联与结构表征视角下的信息隐私研究主题演化研究

情报科学

时序关联与结构表征视角下的信息隐私研究主题演化研究

1. 南京信息工程大学管理工程学院南京 210044,2. 南京信息工程大学自贸区研究院南京 210044

Research on the Topic Evolution of Information Privacy from the Perspective of Temporal Correlation and Structural Representation

1. School of Management Science and Engineering, Nanjing University of Information Science and Technology, Nanjing, 210044
2. Institute of Free Trade Zone, Nanjing University of Information Science and Technology, Nanjing, 210044

摘要/Abstract

摘要：

[目的/意义] 识别信息隐私研究领域的热点主题，梳理主题演化路径。[方法/过程]针对主题识别语义杂乱、主题演化分析片面等问题，提出时序关联与结构表征视角下的主题演化分析方法，首先利用LDA（Latent Dirichlet Allocation）模型识别多时间窗口下的文献主题，进一步运用共词分析绘制语义更为独立的主题凝聚子群。在此基础上，从宏观时序关联维度计算相邻窗口下主题间的相似度，梳理演化路径；从微观结构表征维度，设计主题新颖度、中心性、影响力等计量指标，探寻信息隐私前沿和热点主题的演化变迁。[结果/结论]实证分析结果表明，本文方法可以深度挖掘信息隐私领域研究主题，从宏微观两个维度全面梳理主题的演化路径。研究拓展了主题演化研究的理论视角，有利于探测信息隐私研究的前沿和热点。

关键词:

信息隐私, 时序关联, 结构表征, 主题演化, LDA, 共词分析

Abstract:

[Purpose/Significance] This paper aims to identify the hot topics of information privacy research and summarize the evolution trajectories of research topics.［Method/process］ Considering the research limitations of topic identification and topic evolution, this paper proposed a comprehensive research method of topic evolution from the perspective of temporal correlation and structural representation. 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 macro 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.

Key words:

information privacy, temporal correlation; structural representation; topic evolution; LDA, co-word analysis

朱光潘高枝李凤景.

时序关联与结构表征视角下的信息隐私研究主题演化研究 [J]. 情报科学.

Zhu Guang , Pan Gaozhi Li Fengjing .

Research on the Topic Evolution of Information Privacy from the Perspective of Temporal Correlation and Structural Representation [J]. Information Science.