A FAKE NEWS CLASSIFICATION FRAMEWORK: APPLICATION ON IMMIGRATION CASES

In this study we conduct a systematic literature review with the aim of pointing out the characteristics and types of fake news. We use them to formulate a framework to facilitate classification of fake news instances. Using the classification framework of fake news, we analyse 59 different fake news cases regarding immigration. The research team provided a proof of concept of applicability of the proposed framework for categorising immigration fake news cases. Towards this direction, machine learning algorithms were employed to identify association rules among the classification facets of our framework. The findings of the research study show that a number of these rules can be used in order to design a semi-automated tool that fills-in some of the characteristics of our framework and infers the rest, thus utilising the extracted rules. Benefits stemming from this work include a proposition of an easy to use framework for fake news classification, and derivation of commonly occurring patterns that demonstrate how fake news, as well as their types, interrelate.