Aprendizagem de máquina com representação do conhecimento via linguagem OWL na detecção automática de comportamentos suspeitos em redes sociais

Abstract

The advancement of Artificial Intelligence (AI) has had a transformative impact on the technological landscape and society at large. Since the inception of the Turing Machine, which marked a pivotal point in this trajectory, the understanding of computation and the capacity of machines to perform logical and algorithmic operations have continuously evolved. With the advent of Machine Learning, the way machines process information has undergone a drastic change, enabling the identification of complex patterns within massive datasets. Furthermore, Knowledge Representation, in conjunction with the use of ontologies, the OWL language, and the NLTK library, provides the opportunity to construct systems capable of detecting suspicious behaviors on social networks. The objective of this study is to enhance the semantic analysis of complex interactions on social networks, with a focus on detecting patterns indicative of potentially harmful or threatening activities. Specifically, the study aims to develop an approach that can contribute to the identification of suspicious behaviors on social media platforms. To achieve this objective, we utilize ontologies and the OWL language to model the knowledge and structure of social interactions. This contributes to a more effective approach in identifying and mitigating harmful behaviors on social networks, thereby aiding in the creation of safer environments for users.