Using Convolutional Neural Networks to Classify Hate-Speech

By Björn Gambäck, Utpal Kumar Sikdar. 2017

549 citations recorded by the source corpus.

The paper introduces a deep learning-based Twitter hate-speech text classification system. The classifier assigns each tweet to one of four predefined categories: racism, sexism, both (racism and sexism) and non-hate-speech. Four Convolutional Neural Network models were trained on resp. character 4-grams, word vectors based on semantic information built using word2vec, randomly generated word vectors, and word vectors combined with character n-grams. The feature set was down-sized in the networks by max-pooling, and a softmax function used to classify tweets. Tested by 10-fold cross-validation, the model based on word2vec embeddings performed best, with higher precision than recall, and a 78.3% F-score.

Subfields: Artificial Intelligence, Information Systems, Natural Language Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Information Systems

In Natural Language Processing

Citations in the corpus

Language Generation Models Can Cause Harm: So What Can We Do About It? An Actionable Survey (2022, methodology): “Neural Detection Toxicity Supervised: (Gambäck and Sikdar, 2017; Pitsilis et al., 2018; d’Sa et al., 2020; Xiang et al., 2021); Semi- and Unsupervised: (Korzeniowski et al.”
Large Scale Crowdsourcing and Characterization of Twitter Abusive Behavior (2018, methodology): “A widely used and available dataset in this ﬁeld, is the one in (Waseem and Hovy 2016), employed by several studies working on hate speech detection such as in (Badjatiya et al. 2017), (Gamb¨ack and Sikdar 2017), (Park and Fung 2017), and (Jha and Mamidi 2017).”
Challenges for Toxic Comment Classification: An In-Depth Error Analysis (2018, background): “…Davidson et al., 2017; Nobata et al., 2016; Kennedy et al., 2017; Samghabadi et al., 2017; Robinson et al., 2018) or the use of (deep) neural networks (Ptaszynski et al., 2017; Pavlopoulos et al., 2017; Badjatiya et al., 2017; Vi-gna et al., 2017; Park and Fung, 2017; Gamb¨ack and Sikdar, 2017).”
Cross-Domain Detection of Abusive Language Online (2018, background): “The best performance is most often attained by deep learning models, the most popular being convolutional neural networks (Gamb¨ack and Sikdar, 2017; Potapova and Gordeev, 2016; Pavlopoulos et al., 2017) and variants of recurrent neural networks (Pavlopoulos et al., 2017; Gao and Huang, 2017;…”
A Unified Deep Learning Architecture for Abuse Detection (2018, methodology): “In [15] authors also use deep learning models to address hate speech on Twitter.”
Improving Hate Speech Detection with Deep Learning Ensembles (2018, methodology): “…use of neural networks and embedding methods is worth exploration by NLP and text mining researchers, as the work of (Badjatiya et al., 2017; Gamb¨ack and Sikdar, 2017; Park and Fung, 2017; Severyn and Moschitti, 2015; Stojanovski et al., 2015; Vosoughi et al., 2016; Yang and Eisenstein,…”
Civil Rephrases Of Toxic Texts With Self-Supervised Transformers (2021, background): “…its strategy may involve shifting the original intent, since “republican” is not a non-offensive synonym of “moron”. et al., 2017; Pavlopoulos et al., 2017; Wulczyn et al., 2017; Gamb¨ack and Sikdar, 2017; Fortuna and Nunes, 2018; Zhang et al., 2018a; Van Hee et al., 2018; Zampieri et al., 2019).”
Examining a hate speech corpus for hate speech detection and popularity prediction (2018, background): “a particular hate speech corpus – a Twitter corpus collected by Waseem and Hovy (2016), which has been gaining traction as a resource for training hate speech detection models (Waseem and Hovy, 2016; Gamback and Utpal, 2017; Park¨ and Fung, 2017) – and analyse it critically to better understand its ”
A Review of Standard Text Classification Practices for Multi-label Toxicity Identification of Online Content (2018, background): “Deep learning models with word embeddings as text representations are state-of-the-art text classiﬁcation solutions that show effectiveness in many tasks such as sentiment analysis and the detection of hate speech (Gamb¨ack and Sikdar, 2017).”
Effective hate-speech detection in Twitter data using recurrent neural networks (2018, methodology): “Gambäck & Sikdar [9] employed pre-trained CNN vectors in an effort to predict four classes, and finally achieving slightly higher F-score than character n-grams.”
Cyber Hate Classification: 'Othering' Language And Paragraph Embedding (2018, methodology): “Word vector extraction was also applied to tweets for cyberhate classification by [12], who built four”
Towards Agile Text Classifiers for Everyone (2023, background): “The underlying models for detecting these attributes have also grown in sophistication, from linear models such as support vector machines (MacAvaney et al., 2019) to more modern approaches that rely on deep learning architectures like CNNs (Gambäck and Sikdar, 2017) and, more recently, transformers”
Studying Generalisability across Abusive Language Detection Datasets (2019, background): “Deep Neural networks such as Convolutional Neural Networks, CNN (Zhang et al., 2018; Gamb¨ack and Sikdar, 2017) and variations of Recurrent Neural Networks, RNN (Pitsilis et al., 2018; Gao and Huang, 2017) have seen widespread success, regularly obtaining state-of-the-art results on various…”