Hamdy Mubarak

Abubakr Mohamed, Hamdy Mubarak. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing. 2025.

Hamdy Mubarak, Abubakr Mohamed, Majd Hawasly. Findings of the Association for Computational Linguistics: EMNLP 2025. 2025.

Majd Hawasly, Tasnim Mohiuddin, Hamdy Mubarak, Sabri Boughorbel. Proceedings of The Third Arabic Natural Language Processing Conference. 2025.

Hamdy Mubarak, Rana Malhas, Watheq Mansour, Abubakr Mohamed, Mahmoud Fawzi, Majd Hawasly, Tamer Elsayed, Kareem Mohamed Darwish, Walid Magdy. Proceedings of The Third Arabic Natural Language Processing Conference: Shared Tasks. 2025.

This paper presents a novel Dialectal Sound and Vowelization Recovery framework, designed to recognize borrowed and dialectal sounds within phonologically diverse and dialect-rich languages, that extends beyond its standard orthographic sound sets. The proposed framework utilized a quantized sequence of input with(out) continuous pretrained self-supervised representation. We show the efficacy of the pipeline using limited data for Arabic, a dialect-rich language containing more than 22 major dialects. Phonetically correct transcribed speech resources for dialectal Arabic are scarce. Therefore, we introduce ArabVoice15, a first-of-its-kind, curated test set featuring 5 hours of dialectal speech across 15 Arab countries, with phonetically accurate transcriptions, including borrowed and dialect-specific sounds. We described in detail the annotation guideline along with the analysis of the dialectal confusion pairs. Our extensive evaluation includes both subjective -- human perception tests and objective measures. Our empirical results, reported with three test sets, show that with only one and half hours of training data, our model improve character error rate by ~ 7\% in ArabVoice15 compared to the baseline.

Yassine El Kheir, Hamdy Mubarak, Ahmed Ali, Shammur Chowdhury. Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2024.

Samir Abdaljalil, Hamdy Mubarak. Proceedings of The Second Arabic Natural Language Processing Conference. 2024.

We present an overview of the ArAIEval shared task, organized as part of the first ArabicNLP 2023 conference co-located with EMNLP 2023. ArAIEval offers two tasks over Arabic text: (i) persuasion technique detection, focusing on identifying persuasion techniques in tweets and news articles, and (ii) disinformation detection in binary and multiclass setups over tweets. A total of 20 teams participated in the final evaluation phase, with 14 and 16 teams participating in Tasks 1 and 2, respectively. Across both tasks, we observed that fine-tuning transformer models such as AraBERT was at the core of the majority of the participating systems. We provide a description of the task setup, including a description of the dataset construction and the evaluation setup. We further give a brief overview of the participating systems. All datasets and evaluation scripts from the shared task are released to the research community. (https://araieval.gitlab.io/) We hope this will enable further research on these important tasks in Arabic.

Social media platforms empower us in several ways, from information dissemination to consumption. While these platforms are useful in promoting citizen journalism, public awareness, etc., they have misuse potential. Malicious users use them to disseminate hate speech, offensive content, rumor, etc. to promote social and political agendas or to harm individuals, entities, and organizations. Oftentimes, general users unconsciously share information without verifying it or unintentionally post harmful messages. Some of such content often gets deleted either by the platform due to the violation of terms and policies or by users themselves for different reasons, e.g., regret. There is a wide range of studies in characterizing, understanding, and predicting deleted content. However, studies that aim to identify the fine-grained reasons (e.g., posts are offensive, hate speech, or no identifiable reason) behind deleted content are limited. In this study, we address an existing gap by identifying and categorizing deleted tweets, especially within the Arabic context. We label them based on fine-grained disinformation categories. We have curated a dataset of 40K tweets, annotated with both coarse and fine-grained labels. Following this, we designed models to predict the likelihood of tweets being deleted and to identify the potential reasons for their deletion. Our experiments, conducted using a variety of classic and transformer models, indicate that performance surpasses the majority baseline (e.g., 25% absolute improvement for fine-grained labels). We believe that such models can assist in moderating social media posts even before they are published.

We introduce a generic, language-independent method to collect a large percentage of offensive and hate tweets regardless of their topics or genres. We harness the extralinguistic information embedded in the emojis to collect a large number of offensive tweets. We apply the proposed method on Arabic tweets and compare it with English tweets—analyzing key cultural differences. We observed a constant usage of these emojis to represent offensiveness throughout different timespans on Twitter. We manually annotate and publicly release the largest Arabic dataset for offensive, fine-grained hate speech, vulgar, and violence content. Furthermore, we benchmark the dataset for detecting offensiveness and hate speech using different transformer architectures and perform in-depth linguistic analysis. We evaluate our models on external datasets—a Twitter dataset collected using a completely different method, and a multi-platform dataset containing comments from Twitter, YouTube, and Facebook, for assessing generalization capability. Competitive results on these datasets suggest that the data collected using our method capture universal characteristics of offensive language. Our findings also highlight the common words used in offensive communications, common targets for hate speech, specific patterns in violence tweets, and pinpoint common classification errors that can be attributed to limitations of NLP models. We observe that even state-of-the-art transformer models may fail to take into account culture, background, and context or understand nuances present in real-world data such as sarcasm.