Untitled

Abstract

Abuse on the Internet represents a significant societal problem of our time. Previous research on automated abusive language detection in Twitter has shown that communitybased profiling of users is a promising technique for this task. However, existing approaches only capture shallow properties of online communities by modeling followerfollowing relationships. In contrast, working with graph convolutional networks (GCNs), we present the first approach that captures not only the structure of online communities but also the linguistic behavior of the users within them. We show that such a heterogeneous graph-structured modeling of communities significantly advances the current state of the art in abusive language detection.

Multimodal Hate Speech Detection in Greek Social Media

“…In this paper, we follow and extend the work of Reference [3][4][5] on anti-refugee and anti-migrant hate speech detection. We apply hate speech detection to Greek and enrich this with a multimodal approach, in order to take into account hateful content that does not necessarily carry textual streams.…”

“…They apply a series of models and report best precision, recall, and f1-score of 0.91, 0.90, and 0.90, respectively. Mishra et al [5] use graph convolutional networks to attack the problem, utilising social graph information as part of the model. Waseem and Hovy focus on data collection [3] and on linguistic features that improve quality, while Waseem [4] provides a list of criteria for the annotation process of hate speech.…”

“…While there have been some publicly available datasets of labelled hateful tweets, in the form of tweet IDs (References [3,4,26]), as Mishra observes [5], many of these tweets have been deleted and abusive users have been suspended due to violations of Twitter's policy. Thus, the available datasets are no longer valid for baselines and comparisons and can only be partially used as additional data for model training.…”

Joint Modelling of Emotion and Abusive Language Detection

“…Recently, approaches to abuse detection have moved towards more complex models that utilize auxiliary knowledge in addition to the abuse-annotated data. For instance, Mishra et al (2018aMishra et al ( , 2019a) used community-based author information as features in their classifiers with promising results. Founta et al (2019) used transfer learning to fine-tune features from the author metadata network to improve abuse detection.…”

“…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. However, all of the existing approaches have focused on modelling the linguistic properties of the comments or the meta-data about the users.…”