Us and them: identifying cyber hate on Twitter across multiple protected characteristics

By P. Burnap, Matthew Leighton Williams. 2016

399 citations recorded by the source corpus.

Hateful and antagonistic content published and propagated via the World Wide Web has the potential to cause harm and suffering on an individual basis, and lead to social tension and disorder beyond cyber space. Despite new legislation aimed at prosecuting those who misuse new forms of communication to post threatening, harassing, or grossly offensive language - or cyber hate - and the fact large social media companies have committed to protecting their users from harm, it goes largely unpunished due to difficulties in policing online public spaces. To support the automatic detection of cyber hate online, specifically on Twitter, we build multiple individual models to classify cyber hate for a range of protected characteristics including race, disability and sexual orientation. We use text parsing to extract typed dependencies, which represent syntactic and grammatical relationships between words, and are shown to capture ‘othering’ language - consistently improving machine classification for different types of cyber hate beyond the use of a Bag of Words and known hateful terms. Furthermore, we build a data-driven blended model of cyber hate to improve classification where more than one protected characteristic may be attacked (e.g. race and sexual orientation), contributing to the nascent study of intersectionality in hate crime.

Subfields: Artificial Intelligence, Communication, Law, Natural Language Processing, Safety, Robustness & Interpretability

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 Communication

In Law

In Natural Language Processing

In Safety, Robustness & Interpretability

Citations in the corpus

Language Generation Models Can Cause Harm: So What Can We Do About It? An Actionable Survey (2022, background): “, 2012; Burnap and Williams, 2015; Liu and Forss, 2015); n-gram features (Chen et al., 2012; Waseem and Hovy, 2016; Nobata et al., 2016; Xu et al., 2012; Burnap and Williams, 2016) Misinformation Word-Level features (Zhao et al.”
A Survey on Hate Speech Detection using Natural Language Processing (2017, methodology): “Dependency relationships are also employed in the feature set from Gitari et al. (2015), Burnap and Williams (2015), Burnap and Williams (2016) and Nobata et al. (2016).”
Challenges and frontiers in abusive content detection (2019, background): “Researchers have also addressed numerous tasks beyond binary abusive content classification, including identifying the target of abuse and its strength as well as automatically moderating content (Burnap & Williams, 2016; Davidson, Warmsley, Macy, & Weber, 2017; Santos, Melnyk, & Padhi, 2018).”
Challenges for Toxic Comment Classification: An In-Depth Error Analysis (2018, background): “, 2018), related research includes the investigation of hate speech (Badjatiya et al., 2017; Burnap and Williams, 2016; Davidson et al., 2017; Gambäck and Sikdar, 2017; Njagi et al., 2015; Schmidt and Wiegand, 2017; Vigna et al., 2017; Warner and Hirschberg, 2012), online harassment (Yin and Davison”
HateXplain: A Benchmark Dataset for Explainable Hate Speech Detection (2020, methodology): “To tackle this, various methods have been proposed for hate speech detection (Burnap and Williams 2016; Ribeiro et al. 2018; Zhang, Robinson, and Tepper 2018; Qian et al. 2018a).”
Like Trainer, Like Bot? Inheritance of Bias in Algorithmic Content Moderation (2017, background): “…logics for automated and semi-automated moderation exist, including the promotion of ‘quality’ comments online [13], the ﬂagging of hate speech or bullying [6, 17, 29, 34], or the maintenance of imagined ‘networked publics’ that ‘allow people to gather for social, cultural, and civic purposes’ [5].”
A Framework for the Computational Linguistic Analysis of Dehumanization (2020, neutral): “, 2019), and “othering” language (Burnap and Williams, 2016; Alorainy et al., 2019).”
Human-Machine Collaboration Approaches to Build a Dialogue Dataset for Hate Speech Countering (2022, methodology): “Many benchmarks for automatic HS detection are currently available (Mathew et al., 2021; Cao et al., 2020; Kumar et al., 2018; Hos-seinmardi et al., 2015; Waseem, 2016; Burnap and Williams, 2016).”
Directions in Abusive Language Training Data: Garbage In, Garbage Out (2020, background): “abusive content, which can be deeply shocking, and some training datasets from highly cited publications have not been made publicly available [22, 82, 108].”
Characterizing Linguistic Attributes for Automatic Classification of Intent Based Racist/Radicalized Posts on Tumblr Micro-Blogging Website (2017, background): “Topic Modeling: The existing literature shows that there has been a lot of work in the area of mining user generated content on social media related to offensive speech [13], racism and radicalization [14][15].”
A Survey on Automatic Detection of Hate Speech in Text (2018, background): “Automated techniques aim to programmatically classify text as hate speech, making its detection easier and faster for the ones that have the responsibility to protect the public [9, 65].”
Detecting weak and strong Islamophobic hate speech on social media (2018, background): “It also includes some tweets from before 2017 which are made available by Twitter’s Rest API.”
COLD: Annotation scheme and evaluation data set for complex offensive language in English (2020, background): “Recent computational work has investigated the use of distancing-related features for hate speech classification (Alorainy et al., 2019; Burnap & Williams, 2016), but to our knowledge there is no previous data set which codes for the presence of distancing.”
Hierarchical CVAE for Fine-Grained Hate Speech Classification (2018, background): “Waseem and Hovy (2016), Burnap and Williams (2016), and Davidson et al. (2017) focus on three-way classification.”
Abusive Language Detection using Syntactic Dependency Graphs (2020, background): “On the other hand, Burnap and Williams (2016); Alorainy et al. (2019) proposed computational models using the dependency labels as features and reported signiﬁcant gains over the bag-of-words features.”
Hate Speech Detection: A Solved Problem? The Challenging Case of Long Tail on Twitter (2018, background): “Unfortunately, only a very limited number of work has reported performance on a per-class basis [4, 30].”
Exploring the Role of Grammar and Word Choice in Bias Toward African American English (AAE) in Hate Speech Classification (2022, background): “These include, simple surface features (such as word and character n-grams) [4, 5, 9, 27, 40, 42], word generalization [11, 27], sentiment analysis [40], lexical resources”