Virtual Identity Knowledge Graph Construction Research To Network Threats

This paper focus on three main problems about building knowledge graph for virtual identity: data acquisition, data analysis and consolidation, data storage. Firstly, we propose four methods to get virtual identity information, including 1)collecting leaked user registration informations from web sites, 2)directly crawling user profiles from designated websites, 3)crawling virtual identity from forum messages and blog comments based on the search engine, 4)downloading excel spreadsheet which contains personal information of virtual identity based on search engine. Secondly, this paper makes register mailbox, the user name associated with the registration site as correlation factors to combine the virtual identities which belongs to the same person. In addition, for the demand of high efficiency of database access, data analysis, data consolidation and store analysis results with complex structure, we use MongoDB for storing virtual identity data. Thirdly, we build a prototype system about knowledge graph for virtual identity to validate the data collection, data analysis method, as well as the effectiveness of the way data is stored.