基于网络账号的虚拟社交平台团体关系发现

Group Relationship Discovery of Virtual Social Platforms Based on Network and Media Accounts

设计虚拟社交平台团体关系发现,旨在从个人网络轨迹及舆论热评数据中获取用户的网络社交团体关系,并由此反映跨地区、种族人群间复杂的关联关系。首先,根据网络社团发现需求提出了一种虚拟社交平台团体关系发现闭环结构,设计了模型的输入输出数据结构及信息流程图,以满足网络平台虚拟社团测试集的分类需求;然后,基于多关系社团网络中社区结构检测(CSDM)聚类算法对虚拟账号的活动信息进行理论分析,抽取活跃用户的网络行为轨迹规律;最后,通过基于机器学习的社区检测方法(MLCDM)实现测试集中虚拟网络社团属性挖掘,并给出仿真试验结果分析。

Group relationship discovery of virtual social platforms aims to obtain the relationship of tar-get network social groups from personal network track and public opinion data,and reflect complex re-lationships across regions and ethnic groups.Firstly,a closed-loop structure of group relationship dis-covery of virtual social platforms based on the network community discovery requirements is pro-posed.The structure of input and output data,and information flow chart of the model are designed to meet the classification requirements of network platform virtual community test sets.Then,based on community structure detection in multi-relationships social networks(CSDM)clustering algorithm,the activity information of virtual accounts is analyzed theoretically,and the network behavior trajecto-ry rules of active users are extracted.Finally,machine learning community detection method(MLCDM)is used to mine important community attributes of virtual network in test sets,and analy-sis on simulation results is given.