Threat intelligence analysis based on multi-dimensional feature topic clustering and knowledge graph

In the Internet era, network security experts need to collect and analyze security threats from open source security information in real time, and convert them into data containing high-value information to help security engineers respond quickly or prevent potential network threats. Aiming at massive security information, this paper designs and implements a threat intelligence analysis system based on multi-dimensional feature topic clustering and knowledge graph. Taking the keywords based on TextRank, topic words based on LDA and entity words based on named entity recognition as features, a topic clustering model based on hierarchical clustering is proposed. The knowledge graph is generated based on semantic role annotation to quickly understand the threat information, causality and time dependencies in the topic.