My research sits at the intersection of artificial intelligence, language, and society, with a regional focus on the Middle East and North Africa. Recent work centers on large language models, generative AI, and their social impact, alongside long-standing projects on Arabic language resources and computational social science.
Safety evaluation for under-resourced languages, cultural adaptation of large language models, cognitive alignment, bias auditing, and responsible multilingual LLM development. Active projects span Arabic dialects, Albanian, Kazakh, and Chinese safety evaluation.
Corpus construction, morphological and syntactic annotation, dialectal coverage, and evaluation for Arabic. Long-standing work on open resources for Modern Standard Arabic, Gulf, Levantine, Egyptian, and Maghrebi varieties.
Detection and mitigation of online hate speech, offensive language, and discrimination in Arabic content. Corpus development (ADHAR, MARASTA), shared task organization, and cross-platform studies.
Co-organization of the CLEF CheckThat! Lab, ArAIEval shared tasks, and sustained work on propaganda detection, claim verification, and author profiling for fake-news spreaders.
Computational analysis of religious and historical texts, including Hadith narrator chains, Arabic women writing, and multimedial education tools for Qatari heritage.
Digital mental health platforms for the Arab world, corpus-based studies of online medical discourse, and bilingual mental health NLP benchmarking.
Social Media Analytics and Computational Social Science
Large-scale analysis of Arabic social media for public health, political discourse, societal cohesion, and mental health research. Work spans Twitter, Facebook, YouTube, and cross-platform observational studies.