A Platform Agnostic Dual-Strand Hate Speech Detector

Abstract

Hate speech detectors must be applicable across a multitude of services and platforms, and there is hence a need for detection approaches that do not depend on any information specific to a given platform. For instance, the information stored about the text's author may differ between services, and so using such data would reduce a system's general applicability. The paper thus focuses on using exclusively text-based input in the detection, in an optimised architecture combining Convolutional Neural Networks and Long Short-Term Memory-networks. The hate speech detector merges two strands with character ngrams and word embeddings to produce the final classification, and is shown to outperform comparable previous approaches.

Towards generalisable hate speech detection: a review on obstacles and solutions

“…Three studies compared character-level, word-level, and hybrid (both character- and word-level) CNNs, but drew completely different conclusions. Park (2018) and Meyer & Gambäck (2019) found hybrid and character CNN to perform best respectively. Probably most surprisingly, Lee, Yoon & Jung (2018) observed that word and hybrid CNNs outperformed character CNN to similar extents, with all CNNs performing worse than character n-gram logistic regression.…”

Towards multidomain and multilingual abusive language detection: a survey

“…Their argument for adopting the traditional approach was to provide better explainability of the knowledge transfer between domains. Some other studies adopted several neural-based models, including convolutional neural networks (CNN) [75,141], long short-term memory (LSTM) [8,75,92,94,145], bidirectional LSTM (Bi-LSTM) [115], and gated recurrent unit (GRU) [27]. The most recent works focus more on investigating transferability or generalizability of stateof-the-art transformer-based models such as Bidirectional Encoder Representations from Transformers (BERT) [19,48,66,79,83,90,92,134] and its variant like RoBERTa [48] in the cross-domain abusive language detection task.…”

“…[134] Neural model This study proposed several LSTM-based models that only focuses on using text information (char n-grams and word embedding) representation for building platform-agnostic hate speech detector, but they did not conduct any cross or multidomain experiment to evaluate their model. [75] Transformer based Experimented with a BERT-based classifier and topic modeling approach, which show that removing domain-specific instances improve the model's out-domain performance [83] Neural based Proposed several representations including target, content, and linguistic behavior and used cross attention gate flow to refine these representations, providing better domain-transfer knowledge.…”

Studying Generalisability across Abusive Language Detection Datasets

“…For the most part, word-level n-grams have been highly predictive, with other linguistic features such as part-of-speech tags (Xu et al, 2012;Davidson et al, 2017) and sentiment score (Van Hee et al, 2015;Davidson et al, 2017) providing slight improvements. Due to their ability to perform better in an online setting where spelling errors and adversarial behaviour are commonplace, character-level features have been endorsed , and also shown to often be superior to word-level information for this task (Meyer and Gambäck, 2019). Metadata about users have also been used as features: Waseem and Hovy (2016) claim gender information leads to improved performance, while Unsvåg and Gambäck (2018) report user-network data to be more important.…”