Researchers have reviewed the current state of hadith computational science, which involves using large language models to analyze and interpret Islamic texts. They found that while progress has been made in areas such as data resources and segmentation tasks, there are still significant challenges to overcome, including limited reproducibility and sparse expert validation. The authors argue that hadith computation should be viewed as an evidence infrastructure problem rather than just a model performance issue, requiring knowledge integration and provenance. They propose a research agenda to address these gaps and make the field more useful to Islamic scholarship. --- Why it matters: This review matters because it highlights the limitations of current methods in hadith computational science and proposes a new approach that could lead to more robust and reliable results. This is important for researchers who want to use AI to analyze and interpret Islamic texts, as well as for scholars who rely on these texts for their research. Source: https://arxiv.org/abs/2608.20364