Moheet Corpus
SADIQOON TECHNOLOGIES LTD

METHODOLOGY PAPER · NOVEMBER 2026

Moheet Corpus: A Methodology for Building a Citation-Grade Semantic Corpus of Arabic and Persian Islamic Scholarship

From library file to cited passage: structure-preserving ingestion, hybrid retrieval, an evidence-bound knowledge graph, and measured evaluation

A methodology paper on how Moheet Corpus was built over the Ahl al-Bayt Library in Arabic and Persian: the cited passage as unit of construction, structure-preserving ingestion that keeps editors' footnotes in place, vector representation and hybrid retrieval, a knowledge graph that records no relation without literal evidence, measured figures, and a commitment to open release. By Shady Ali.

Open the paper (PDF)
12 pages · 138 KB

DOI: 10.5281/zenodo.23865623 · CC BY 4.0

Cite as:
Ali, Shady. Moheet Corpus: A Methodology for Building a Citation-Grade Semantic Corpus of Arabic and Persian Islamic Scholarship. Zenodo, 2026. https://doi.org/10.5281/zenodo.23865623

Also available: العربية · Introductory paper: The Moheet Corpus: A Shift in the Architecture of Shi’i Islamic Knowledge

← Back to the Moheet Corpus page