Multi contextual learning of disinformation on social media

Abstract

Not only does disinformation research have to deal with multiple contexts, but it is also inherently multidisciplinary because of the humans and society involved. This makes it a vast interdisciplinary research canvas. Motivated by this, this research focused on objectives of two types, i.e., macro challenges of understanding the complexities of the big picture and relatively micro challenges of implementing machine learning approaches in disinformation research. The first research contribution addresses a macro challenge of obtaining a systemic view by identifying the commonalities across the three main vertical silos of disinformation research, i.e., fake news, rumor, and propaganda. It does this from the perspective of multi-contextuality that affects all these areas. newline The second research contribution targets an interdisciplinary macro challenge in online social movements that are common brewing grounds of disinformation campaigns in today s world. The research undertaken makes two specific contributions. First, it offers a novel socio-psychological analysis of the organizers and participants in such social movements. Examining moral framing and Basic Human Needs extends the why s of social movements and brings a deeper user context so far overlooked by disinformation research. Second, it provides an analytical grid explaining the how and spanning the technical machinations of mobilization and counter-mobilizations that have not been attempted so far except for a partial attempt for the Black Lives Matter movement. The framework proposed can be adapted to analyze any future online social movement. newline Disinformation has a deeper context, i.e., the human mind, which is crucial in deciding participation and behavior. So far, linguistic markers have been used as a proxy for the mind, and direct modeling of the human psychology at play in disinformation has not been attempted. The third research contribution extends the work done in the second contribution and builds on the psychological context of disinformatio