Multi-Dimensional NLP Analysis of the Quran: Structure, Semantics, and Emotion across Qirāʾāt

Abstract

This study investigates the Quranic text through a multi-dimensional Natural Language Processing (NLP) framework that integrates structural, semantic, and emotional analysis. Drawing on both Arabic and translated corpora, we apply a combination of traditional statistical techniques (e.g., TF-IDF) and transformer-based models (e.g., AraBERT) to examine thematic distributions, linguistic variation, and affective tone across Surahs. Our analysis incorporates tokenization, frequency analysis, part-of-speech tagging, and emotion classification, supported by visualizations such as bar charts and word clouds. Part-of-speech tagging refers to labeling each word in the Quranic text with its grammatical role (e.g., noun, verb, adjective), which helps clarify the structure and meaning of verses. Results reveal distinct structural and emotional profiles between Meccan and Medinan Surahs, with shorter chapters exhibiting denser emotional content, and longer ones displaying greater thematic diversity. The inclusion of Qirāʾāt perspectives further highlights meaningful canonical variation often overlooked in computational studies. This work contributes to the emerging field of Quranic computational linguistics by offering an integrative approach that bridges traditional exegesis with modern NLP and lays the groundwork for future applications in tafsir automation, recitation analysis, and discourse modeling.