Islam and Data Science Roundup

In “Technical vs Cultural: Evaluating LLMs in Arabic” (Open Review), Ahmad A. Rushdi (Stanford University) “pilot evaluation framework for language models in Arabic, revealing nuanced performance patterns across technical and cultural dimensions. We evaluate five prominent models—Arabic-specialized systems (Fanar, Falcon 3) and frontier models (Claude Opus, GPT-5, Llama)—across a small set of 45 prompts spanning… CONTINUE READING

In “Comparative Analysis of the Risk of Hadith Errors in Question-Answering Systems Based on Large Language Models” (International Journal of Research and Scientific Innovation), Hakkun Elmunsyah (Universitas Negeri Malang) conducts a “comparative analysis of the risk of errors generated by Large Language Model-based Question-Answering systems in answering hadith-related questions. Four LLM models, namely GPT, Gemini,… CONTINUE READING

In “Reimagining Islamic Civilization in the Age of AI: Digital Humanities, Ethical Challenges, and Knowledge Production in the Muslim World” (Islamic Civilization and History), Muhammad Yusuf (Institut Agama Islam Negeri Syekh Nurjati) observes that “the convergence of artificial intelligence with Islamic civilization presents unprecedented challenges to knowledge production, ethical frameworks, and educational practices. Despite producing… CONTINUE READING

In “Mapping Qur’anic Contemplation Research (2001–2025): Trends, Contributions And Future Directions” (Quranica), Ahmad Amirul Ihsan Che’ Zahari (Universiti Sains Malaysia) and Mohd Shukri Hanapi (Universiti Sains Malaysia) examine “the development and trends in Qur’anic contemplation (tadabbur) research based on publications indexed in the Scopus database between 2001 and 2025….Using VOSviewer, R-Studio (Bibliometrix) and Microsoft Excel,… CONTINUE READING

In “Artificial Intelligence and Knowledge Production in Brill’s Encyclopaedias of Islamic Studies: A Statistical and Analytical Investigation” (Contemporary Journal of Social Science Review), Hafiz Mohsin zia Qazi (Muslim Youth University) “critically examines the impact of AI on Islamic encyclopaedic scholarship, with a particular focus on Brill’s Encyclopaedia of Islam as a foundational reference work in… CONTINUE READING

In “AI and Qur’anic Interpretation: Exploring the Ethical and Epistemological Boundaries of Artificial Intelligence in Understanding the Qur’an” (Al-Furqan), Safa Alrumayh (University of Zawia) “critically examines the role of AI in Qur’anic interpretation by addressing its epistemic capabilities, ethical limitations, and claims to interpretive authority. Employing a qualitative, literature-based research design, this study analyzes peer-reviewed… CONTINUE READING

In “OpenITI MAKHZAN: An Open Annotated Dataset of Arabic, Persian, Ottoman Turkish, and Urdu Print and Manuscript Data” (Journal of Open Humanities Data), Jonathan Parkes Allen (University of Maryland) and others introduce OpenITI MAKHZAN, a “large aggregation of Arabic-script ground truth and evaluation data drawn from a wide variety of Persian, Arabic, Ottoman Turkish, and… CONTINUE READING

In “OpenITI MAKHZAN: An Open Annotated Dataset of Arabic, Persian, Ottoman Turkish, and Urdu Print and Manuscript Data” (Journal of Open Humanities Data), Jonathan Parkes Allen (University of Maryland) and others introduce OpenITI MAKHZAN, a “large aggregation of Arabic-script ground truth and evaluation data drawn from a wide variety of Persian, Arabic, Ottoman Turkish, and… CONTINUE READING

In “Sacred or Synthetic? Evaluating LLM Reliability and Abstention for Religious Questions” (arXiv), Farah Atif (Mohamed Bin Zayed University of AI) and others “introduce a novel benchmark FiqhQA focused on the LLM generated Islamic rulings explicitly categorized by the four major Sunni schools of thought, in both Arabic and English. Unlike prior work, which either… CONTINUE READING

In “HATFormer: Historic Handwritten Arabic Text Recognition with Transformers” (arXiv), Adrian Chan (independent scholar) and others “propose HATFormer, a transformer-based encoder-decoder architecture that builds on a state-of-the-art English HTR model. By leveraging the transformer’s attention mechanism, HATFormer captures spatial contextual information to address the intrinsic challenges of Arabic script through differentiating cursive characters, decomposing visual… CONTINUE READING