基于网络嵌入方法的耦合网络节点表示学习  

Coupling network vertex representation learning based on network embedding method

作　　者：韩忠明[1,2] 刘聃 郑晨烨 刘雯 段大高[1] 董健[3] Zhongming HAN;Dan LIU;Chenye ZHENG;Wen LIU;Dagao DUAN;Jian DONG(College of Computer and Information Engineering,Beijing Technology and Business University,Beijing 100048,China;Beijing Key Laboratory of Food Safety and Big Data Technology,Beijing 10004,China;Key Laboratory of Information Network Security Ministry of Public Security,the Third Research Institute of the Ministry of Public Security,Shanghai 200031,China)

机构地区：[1]北京工商大学计算机与信息工程学院,北京100048 [2]食品安全大数据技术北京市重点实验室,北京100048 [3]公安部第三研究所信息网络安全公安部重点实验室,上海200031

出　　处：《中国科学：信息科学》2020年第8期1197-1216,共20页Scientia Sinica(Informationis)

基　　金：国家自然科学基金(批准号:61170112);北京市自然科学基金(批准号:4172016);北京市教委科技计划一般项目(批准号:KM201710011006);公安部重点实验室开放课题资助项目。

摘　　要：网络节点表示学习是网络数据分析挖掘中的一个基础问题,通过学习网络节点表示向量,可以更加精准地对网络节点进行表征.近年来,随着深度学习的发展,嵌入方法在网络节点表示学习方面得到了广泛应用.同时,网络数据在规模、模态等特征方面也有了很大的变化,研究重点从单网络分析挖掘逐渐演变至耦合网络分析挖掘.本文首先分析了嵌入方法在单网络节点表示学习中的研究现状,对比了现有方法的优劣.然后借鉴单网络中嵌入方法的思想,针对耦合网络提出了耦合网络嵌入模型CWCNE.针对耦合网络的特性,改进了嵌入方法中的游走算法,提出了一种网络间带约束的随机游走策略;同时改进了模型的训练方法,使用网络间迭代训练的方式来学习模型参数.最后使用了社交耦合网络、学术耦合网络、影视耦合网络、诗词耦合网络、著作耦合网络等5组数据集验证了CWCNE的有效性.并在社团划分、实体识别、标签分类等任务上取得了良好的结果.Network representation learning is a basic problem in network data analysis.By learning network representation vectors,network vertices can be represented more accurately.With the development of deep learning,embedding methods have been widely used for network vertex representation learning.Providing that network data have changed in terms of their scale and modality,the research focus gradually shifted from single network mining to coupling network mining.This paper first analyzes the research status of embedding methods for single networks and then compares their advantages and disadvantages.Furthermore,the paper presents a model called CWCNE for coupling network embedding.The random walk and training algorithms of the proposed model are improved to adapt to coupling network features.The validity of the proposed model was verified using social,academic,film,poetry,and work coupling network data.Good results were obtained on community detection,entity recognition,and label classification tasks.

关 键 词：网络嵌入 节点向量 耦合网络 表示学习 社团划分 主体识别 标签分类 

分 类 号：O157.5[理学—数学]