FCTNLP: An architecture to fight cyberterrorism with natural language processing

Law Enforcement Agencies (LEA) are everyday more and more concerned about illicit activities that may be found in cyberspace like cybercrimes, cyber espionage, cyberterrorism, cyber warfare, among others. In a cyberterrorism context, Hostile Social Manipulation (HSM) is a strategy that employs different manipulation methods mostly through social media to produce damage to a target state. The efforts to fight cyberterrorism could come along with new technologies that
allow a faster and more effective control of offensive actions. For that reason, this paper proposes an artificial intelligence-based solution that processes posts in social networks using Natural Language Processing (NLP) techniques, applying the following three models: i) Sentiment Model to discriminate between threat
and non-threat publications, ii) Similarity Model to identify suspects with similar intentions and iii) NER model that identifies entities in the text. Finally, the proposal was tested exhaustively to validate its functionality and feasibility, achieving an integrated and simple prototype.