Estado del arte de técnicas de inteligencia artificial que aporten en la ciberseguridad

Abstract

The current research examines artificial intelligence (AI) techniques contributing to cybersecurity through mapping and a systematic literature review covering the period from 2018 to 2023. The PICOC method was employed, and researchrelated questions were formulated. Requirements for selection and exclusion were applied in three filtering stages, resulting in the selection of 30 studies. The taxonomy identified various AI techniques, such as Explainable AI (XAI), BT-AI, machine learning, incremental learning, and deep learning. It was concluded that these techniques play a key role in strengthening cybersecurity, addressing threats such as Denial-of-Service (DoS) attacks, malware, and ransomware. Tools related to AI were examined, including Big Data, IoT, UML, experimental analysis, association rules, and confirmatory factor analysis, highlighting their contribution to protecting against cyber attacks. A variety of risks were addressed, from DoS attacks to protection for database management systems, emphasizing a constant need for innovative strategies to safeguard systems and data in cyberspace.