罚似然图模型与社会网络测量

发布时间：2018-06-19 14:57

本文选题：社会网络测量 + 双模网络　；参考：《社会》2017年02期

【摘要】：随着互联网及智能设备的普及,越来越多的用户行为轨迹和互动数据的获得成为可能并进入社会学研究者的视野。这类行为或互动事件的数据在数据结构上属于社会网络分析方法中常见的双模网络。但传统的社会网络分析所面对的数据规模较小,研究者一般采用矩阵分解、主成分分析等描述性分析方式来对网络子群进行区分或测量。而在大数据的背景下,参与互动的群体规模巨大、群体成员的构成动态变化、事件具有时序特征、事件发生存在异质性等特征,使得传统的分析方法无法有效应对此类数据。近十年来,高维高斯图模型在网络关系探测研究中被广泛应用。本文拟对基于罚似然回归的高斯图模型进行综述。罚似然高斯图模型是一个发展迅速的分析工具,本文并不侧重具体的算法和优化过程,而是就罚似然图模型及其扩展模型对社会科学应用研究可能带来的贡献进行梳理。最后,本文亦对涉及的相关模型及其R软件包进行汇总,以期拓展该方法在社会科学领域的应用。
[Abstract]:With the popularity of the Internet and smart devices, more and more users' behavior trajectories and interactive data become possible and enter the field of view of sociological researchers. The data of this kind of behavior or interactive event belong to the double mode network which is common in the social network analysis method in data structure. However, the traditional social network analysis is faced with a small scale of data, researchers generally use matrix decomposition, principal component analysis and other descriptive analysis methods to distinguish or measure network subgroups. Under the background of big data, the large scale of the group participating in the interaction, the dynamic changes of the composition of the group members, the temporal characteristics of the events and the heterogeneity of the events make the traditional analysis methods unable to deal with this kind of data effectively. In recent ten years, high-dimensional Gao Si graph model has been widely used in the research of network relationship detection. In this paper, the Gausttu model based on penalty likelihood regression is reviewed. Penalty likelihood Gao Si graph model is a rapidly developing tool. This paper does not focus on the specific algorithm and optimization process, but reviews the contribution of penalty likelihood graph model and its extended model to the research of social science application. Finally, the related models and their R software packages are summarized in order to expand the application of this method in the field of social science.
【作者单位】：中国社科院社会发展战略研究院;
【基金】：国家社会科学基金一般项目“基于手机大数据的社会心态研究”(16BSH013)的资助~~
【分类号】：TP393.09

【相似文献】