A Semantic Model for Context-Based Fake News Detection on Social Media

Context-based fake news detection provides means to define and describe a social context for news objects on social media, thereby facilitating detection of fake news through data analysis and patterns recognition. However, while content-based fake news detection has gained popularity with machine learning and NLP techniques, the context-based approach has seen very little exploitation. Therefore, it has become pertinent to significantly explore and integrate other technologies for context-based detection of fake news on social media. With semantic technologies capabilities to provide context-awareness for data, this paper analyses social media context and develops a taxonomy for entities classification. Furthermore, a semantic model is developed to describe classes extracted from the taxonomy towards fully semantically describing concepts, relations, instances, and axioms. The model would enhance fake news detection through semantic annotation for contextual features of news objects and datasets, providing a basis for patterns recognition, analysis, and identification of news articles on social media as either fake or not.