A Transformer-Based Approach to Authorship Attribution in Classical Arabic Texts

Authorship attribution (AA) is a field of natural language processing that aims to attribute text its author. Although the literature includes several studies on Arabic AA in general, applying classical texts has not gained similar attention. This study focuses investigating recent pretrained transformer-based models rarely studied domain with limited research contributions: Islamic law. We adopt an experimental approach investigate AA. Because no dataset been designed specifically for this task, we design and build our own using law digital resources. conduct experiments fine-tuning four models: AraBERT, AraELECTRA, ARBERT, MARBERT. Results indicate task attributing given author, ARBERT AraELECTRA outperform other accuracy 96%. conclude transformer models, fine-tuned legal dataset, show significant results texts.