Machine Learning for Identifying Abusive Content in Text Data

Abstract

The proliferation of social media has created new norms in society. Incidents of abuse, hate, harassment and misogyny are widely spread across social media platforms. With the advancements in machine learning techniques, advanced text mining methods have been developed to analyse text data. Social media data poses additional challenges to these methods due to their nature of short content and the presence of ambiguity, errors and noises in content. In the past decade, machine learning researchers have focused on finding solutions dealing with these challenges. Outcomes of these methods boost the social media monitoring capability and can assist policymakers and governments to focus on key issues. This chapter will review various types of machine learning techniques including the currently popular deep learning methods that can be used in the analysis of social media data for identifying abusive content.