Yazar "Khalid, Faiza" seçeneğine göre listele

Listeleniyor 1 - 1 / 1
  • Küçük Resim
    Öğe
    Deepfakes in cybersecurity: a holistic and foundational ontology
    (Karabük Üniversitesi, Lisansüstü Eğitim Enstitüsü, 2024) Khalid, Faiza; Menemencioğlu, Oğuzhan
    Artificial intelligence-based deepfakes are emerging as a significant cybersecurity threat. With the continuous advancement in artificial intelligence tools and techniques to manipulate media, the threat of deep synthetic multimedia fakes has reached the lives of common individuals and business entities alike. The growing adverse impacts of deepfakes on the privacy and security of individuals and government institutes call for the understanding of the domain by the public and development of advanced defense techniques by the government and private researchers. Semantic web helps make data readable and understandable for both humans and machines via ontologies and knowledge graphs. A comprehensive deepfake domain ontology can help structure knowledge of deepfake attack scenarios for better domain understanding. The knowledge graph representation of deepfake attack events, based on the developed ontology, can help visualize these events and infer essential information related to deepfake detection and prevention. This thesis develops the first deepfake attacks ontology and then performs ontology evaluation through knowledge graph application. For a foundational ontology, authors focused on structuring knowledge related to a deepfake attack, like the vulnerable entity, deepfake creator, attack goal, medium, generation technique, consequences, preventive measures, etc. The ontology defines 19 core classes that represent the deepfake attack scenario along with 28 kinds of relations that describe the relation between these entities. The authors used knowledge engineering methodology with 7 steps, including ontology scope determination, existing ontologies evaluation, and classes, properties, and relations definitions. They additionally, utilized Protégé Desktop and the W3C Web Ontology Development Language for ontology creation, the WIDOCO tool for ontology documentation, and OOPS for ontology validation. The authors developed a small size deepfake events knowledge base to implement knowledge graphs, where the developed ontology defined the nodes and relations. GraphDB, a graph database, was used for knowledge graph implementation. Furthermore, this thesis presents knowledge graphs creation of deepfake attack scenarios. The authors implemented knowledge graphs to evaluate ontology’s effectiveness in helping understand and infer deepfake event data. The authors created Semantic Web Rule Language (SWRL) rules that helped infer additional information from the deepfake attack knowledgebase via knowledge graphs application, such as various ways a particular entity can be affected by a deepfake, mediums used for attacks, and online security measures victims can adopt. The textual data of 35 plus global deepfake events in the context of politics, law, world security, etc. is collected and visualized in the form of knowledge graphs to evaluate information understanding and inference capabilities of the first-of-its-kind foundational deepfake ontology. The developed ontology could be used to promote domain understanding and as a framework to build cybersecurity systems with better knowledge inference (semantic reasoning). The ontology can be extended iteratively with new domain advancements. As a future work, NLP approaches could be adopted to automate domain entity research and deepfake event knowledge base population.