Open AccessUsing Machine Learning to Inductively Learn Semantic Rules
Author(s) -
Mohammed Hussein Jabardi,
Asaad Sabah Hadi
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1804/1/012099
Subject(s) - computer science , ontology , semantic web , world wide web , social semantic web , semantic web stack , population , semantic web rule language , representation (politics) , web ontology language , knowledge representation and reasoning , artificial intelligence , information retrieval , internet privacy , semantic analytics , philosophy , demography , epistemology , sociology , politics , political science , law
The Semantic Web and Machine Learning usually are seen as incompatible approaches toward Artificial Intelligence. A proposal presented for integrating the two paradigms and used data from Twitter regarding legitimate and fake accounts. Online Social Networks (OSN) such as Twitter have become a part of our lives due to their ability to connect peo-ple around the world, share documents, photos, and videos. OSN’s such as Facebook, Twitter and LinkedIn have approximately 500 million users over the world; this massive population of OSN causes different kinds of problems regarding data security and privacy. Unauthorised users infringe on the privacy of legitimate users and abuse names and cre-dentials of victims by creating a fake account. We utilised Machine Learning to inductive-ly learn the rules that distinguished a phoney account from a real one. We then imple-mented those rules in a Web Ontology Language (OWL) ontology using the Semantic Web Rule Language (SWRL). This integration provides the benefits of the data-driven ML approach combined with the explicit knowledge representation and the resulting ease of explanation and maintenance of the Semantic Web paradigm.