Abusive language - or, in another very common terminology, hate speech - and the propagation of harmful stereotypes have unfortunately become commonplace occurrences on various social media platforms, partly due to users’ freedom and anonymity and the lack of regulation provided by these platforms. The sheer volume and often implicit nature of such unwanted content make manual moderation of these user spaces a formidable task. Various scientific communities interested in its at least partial automation have taken up the problem over the past ten years. In particular, Computational Social Science, Natural Language Processing and Computational Linguistics have proposed numerous works to create resources, datasets, and models aimed at automating the task of abusive language detection (henceforth ALD). In fact, we see that ALD has become a research theme in its own right in the field of Natural Language Processing with an abundant literature.

Abusive language (we use this term here as an umbrella term to refer to the various forms of harmful language, such as  toxic, offensive language, hate speech, and stereotypes) is topically focused and each specific manifestation of abusive language targets different vulnerable groups based on characteristics such as gender (misogyny, sexism), ethnicity, race, religion (xenophobia, racism, Islamophobia), sexual orientation (homophobia), and so on. Most automatic ALD approaches cast the problem into a binary classification task but important considerations should be taken into account, in particular: (1) the topical focus or the target-oriented nature of hate speech ; (2) the degree of engagement of users in abusive content (e.g., denunciation, approbation, reporting, neutral attitude) ; (3) the question of stereotypes and dominant ideologies ; (4) the question of linguistic strategies more particularly linked or born with social networks (e.g., emoticons, hashtags). Furthermore, most of the work (resources, classifiers) is developed for English.  

**Topics**

Motivated by the interest of the community in the problem of ALD, we invite papers from  Natural Language Processing, Machine Learning and Computational Social Sciences. We explicitly encourage interdisciplinary submissions (resources, computational methods, and user applications at the interface of linguistics/psychology/socio-linguistics/sociology) but also position papers on the actual state of the art in the field discussing the limitations of the current approaches and directions for future work. The topics covered by the special issue include, but are not limited to:

• Linguistic resources and evaluation: annotation schemes, corpus linguistics studies, new datasets, with a particular interest in French language and/or multilingual resources. In the case of strictly lexical resources: methods for constituting them and coverage, semantic categories retained.
• Formal/Conceptual approaches for ALD as inspired by models in sociology, socio-linguistics and psychology.
• Models and Methods: supervised and unsupervised approaches, including LLMs.
• Role of contextual phenomena, including discourses, extra-linguistic contexts (e.g., cultural aspects).
• Models for cross-lingual and multimodal detection.
• New approaches beyond binary classification: target-oriented ALD, degrees of user engagement, etc.
• Dynamics of online AL in social media, propaganda propagation.
• Bias detection and removal in resource creation, datasets and methods.
• Application of ALD tools in education, social media content moderation, etc.
• Social, legal, and ethical implications of detecting, monitoring and moderating AL. 

IMPORTANT DATES

• Submission deadline: May 31th, 2024
  
• Notification to the authors after first review: July 15th, 2024
  
• Notification to the authors after second review: mid-October, 2024
  
• Publication : January 2025

THE JOURNAL

TAL (Traitement Automatique des Langues / Natural Language Processing) is an international journal published by ATALA (French Association for Natural Language Processing, http://www.atala.org) since 1959 with the support of CNRS (National Centre for Scientific Research). TAL has an electronic mode of publication.