Research On Key Technologies Of Building Sensitive Information Knowledge Base Of Zeronet Anonymous Network

Anonymous network has become the focus and difficulty of cyberspace security research due to its anonymous services and a large number of illegal activities.The ZeroNet network is an emerging and rapidly developing anti-censorship distributed anonymous network.In order to support the security of cyberspace,tracing the roots of illegal activities and characters has become a new requirement for current cyberspace security.This thesis integrates site detection technology,node detection technology,crawler technology,knowledge extraction,knowledge fusion and knowledge processing technology to model the nodes,sites,entities,concepts,attributes and the relationships between them in the ZeroNet network at the information level.Then analyze the potential relationships between sites on this basis.The research content of this topic is of great significance in tracing the roots of anonymous network characters and information.This thesis construct the sensitive information knowledge base of the ZeroNet anonymous network and realize visualization,which facilitate the observation and analysis of the nodes and sites,help discover the people,events and site clusters with sensitive information.The traceability of user characters and information provides a large amount of reliable information basis to further support cyberspace security governance and improve the quality of cyberspace security governance.This thesis first describes the origin of the ZeroNet anonymous network,elaborates the ZeroNet network from the characteristics of the ZeroNet network structure and its network architecture,and briefly introduces the related technologies of knowledge base construction.Then,this thesis describes the key technologies of ZeroNet node and site detection,anonymous web crawler and analysis and knowledge base construction,including knowledge extraction,knowledge fusion and knowledge processing technology at home and abroad.Next,according to the characteristics of the ZeroNet anonymous network,this thesis probes the nodes and sites in the ZeroNet network,obtains the corresponding node and site information,and uses crawler technology to obtain the basic information and unstructured data of the site pages in the ZeroNet network.At the information level,this thesis cleans the acquired data,and then uses deep learning technology to extract key information entities and the entities' relationships in the cleaned unstructured data;at the knowledge level,this thesis uses Knowledge fusion technology to map the entities in the ZeroNet text to the corresponding real Internet entities to eliminate the ambiguity of entities in the content of the ZeroNet site and then link the clear knowledge entities to the entities in the knowledge base.Finally,this thesis integrates the network layer and information layer knowledge of nodes and sites,analyzes the core nodes and site relationships of ZeroNet sites,and builds a relatively complete ZeroNet sensitive information knowledge base.This thesis proposes a named entity recognition algorithm model for ZeroNet anonymous network,and integrates three types of anonymous network idiom dictionary features into the algorithm model,which greatly improves the recognition effect of the named entity recognition model on anonymous network site text idioms.At the same time,the network layer and information layer of the ZeroNet anonymous network are merged into the knowledge base.What's more,site relationship analysis algorithm that merges the network layer and information layer information is proposed.In addition,the thesis further realizes the function of visualizing ZeroNet anonymous network.Finally,this thesis summarizes the difficulties and challenges faced by ZeroNet's key technology of sensitive information knowledge base construction,and looks forward to the follow-up work.