PRTIRG: A Knowledge Graph for People-Readable Threat Intelligence Recommendation

People-Readable Threat Intelligence (PRTI) recommender Systems aim to address the problem of information explosion of PRTIs and make personalized recommendation for users. In general, PRTI is highly condensed, and consists of security items, network entities and emerging hacker organizations, attacks, etc. PRTI may also contain many Machine-Readable Threat Intelligence (MRTI). However, existing methods are unaware of such external knowledge and cannot fully discover latent knowledge-level connections among PRTIs. Under this scenario, the existing generic knowledge graphs will introduce too much noise and can not consider the entity relationship in terms of the attack chain. To solve the problems above, in this paper, we propose a knowledge graph for People-Readable Threat Intelligence recommendation (PRTIRG) and incorporates knowledge graph representation into PRTI recommender system for click-through prediction. The key components of PRTIRG are the denoising entity extraction module and the knowledge-aware long short-term memory neural network (KLSTM). Through extensive experiments on real-world datasets, we demonstrate that the PRTIRG is more effective and accurate than baselines.