A Decade in Computational Quranic Studies (2016-2025): A Systematic Survey of the Evolution from Statistical Methods to Large Language Models

Abstract

The intersection of artificial intelligence and sacred scripture represents a critical yet underexplored frontier in the digital humanities. This paper presents the first decade-spanning, systematic survey of computational Qur’anic studies (2016–2025), tracing the field’s evolution from fragmented, ontology-based prototypes to a coherent, methodologically mature research discipline. Through a rigorous, multi-phase systematic review of 112 peer-reviewed studies, we map a transformative trajectory across three eras: a Foundational Era (2016–2019) focused on knowledge representation; a Transformer Turn (2020–2023) driven by community benchmarks and Arabic transformer models; and an Acceleration Era (2024–2025) marked by the integration of Large Language Models (LLMs) and Retrieval-Augmented Generation (RAG). Our longitudinal analysis uncovers a clear paradigm shift—from rule-based and statistical methods toward deep architectures explicitly engineered for ethical and theological fidelity. While remarkable advances have been achieved in Qur’anic Question Answering, Verse Authentication, and Recitation Analysis, the field still faces major gaps in multilingual applications, computational modeling of canonical recitation traditions (Qira¯’a¯t), and the deep integration of classical tafs¯ır scholarship. By introducing a structured taxonomy, a diachronic synthesis of research trends, and a strategic roadmap for the next decade, this survey reframes computational Qur’anic studies as both a case study and a catalyst for innovation in responsible, explainable, and theologically grounded AI.