Joint Modelling of Emotion and Abusive Language Detection

Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 4270–4279 July 5 - 10, 2020. c⃝2020 Association for Computational Linguistics 4270 Joint Modelling of Emotion and Abusive Language Detection Santhosh Rajamanickam ILLC, University of Amsterdam rajamanickamsanthosh@gmail.com Pushkar Mishra Facebook AI pushkarmishra@fb.com Helen Yannakoudakis Dept.of Informatics, King’s College London helen.yannakoudakis@kcl.ac.uk Ekaterina Shutova ILLC, University of Amsterdam e.shutova@uva.nl Abstract The rise of online communication platforms has been accompanied by some undesirable effects, such as the proliferation of aggres- sive and abusive behaviour online. Aiming to tackle this problem, the natural language pro- cessing (NLP) community has experimented with a range of techniques for abuse detec- tion. While achieving substantial success, these methods have so far only focused on modelling the linguistic properties of the com- ments and the online communities of users, disregarding the emotional state of the users and how this might affect their language. The latter is, however, inextricably linked to abu- sive behaviour. In this paper, we present the first joint model of emotion and abusive lan- guage detection, experimenting in a multi-task learning framework that allows one task to in- form the other. Our results demonstrate that incorporating affective features leads to signif- icant improvements in abuse detection perfor- mance across datasets. 1 Introduction Aggressive and abusive behaviour online can lead to severe psychological consequences for its vic- tims (Munro, 2011). This stresses the need for automated techniques for abusive language detec- tion, a problem that has recently gained a great deal of interest in the natural language processing com- munity. The term abuse refers collectively to all forms of expression that vilify or offend an individ- ual or a group, including racism, sexism, personal attacks, harassment, cyber-bullying, and many oth- ers. Much of the recent research has focused on detecting explicit abuse, that comes in the form of expletives, derogatory words or threats, with sub- stantial success (Mishra et al., 2019b). However, abuse can also be expressed in more implicit and subtle ways, for instance, through the use of am- biguous terms and figurative language, which has proved more challenging to identify. The NLP community has experimented with a range of techniques for abuse detection, such as recurrent and convolutional neural networks (Pavlopoulos et al., 2017; Park and Fung, 2017; Wang, 2018), character-based models (Nobata et al., 2016) and graph-based learning methods (Mishra et al., 2018a; Aglionby et al., 2019; Mishra et al., 2019a), obtaining promising results. How- ever, all of the existing approaches have focused on modelling the linguistic properties of the comments or the meta-data about the users. On the other hand, abusive language and behaviour are also inextri- cably linked to the emotional and psychological state of the speaker (Patrick, 1901), which is re- flected in the affective characteristics of their lan- guage (Mabry, 1974). In this paper, we propose to model these two phenomena jointly and present the first abusive language detection method that incor- porates affective features via a multitask learning (MTL) paradigm. MTL (Caruana, 1997) allows two …

Content truncated to 500 words