islamic-scholars / README.md
sahsimanevi's picture
Authority file for the scholars whose work the corpus carries — 2,420 rows
c417fbf verified
|
Raw
History Blame Contribute Delete
2.33 kB
---
license: cc-by-nc-sa-4.0
pretty_name: "islamlab — Islamic Scholars Authority File"
language:
- ar
task_categories:
- text-retrieval
- question-answering
- text-generation
size_categories:
- 100K<n<1M
tags:
- islam
- arabic
- authority-file
- prosopography
- bibliography
- classical-arabic
configs:
- config_name: default
data_files:
- split: train
path: train-*.parquet
---
# islamlab — Islamic Scholars Authority File
One row per scholar for the **2,420 authors** whose work the corpus
carries: the name, the death year in both calendars, the school he wrote in
where the sources state one, the sciences he worked across, and the titles we
hold.
Small table, but it is the one you need first: everything else in islamlab is
keyed by work, and this is what turns a work into a person.
Spread by Hijri century — 1c: 53 · 2c: 49 · 3c: 220 · 4c: 274 · 5c: 246 · 6c: 213 · 7c: 237 · 8c: 331 · 9c: 206 · 10c: 152 · 11c: 100 · 12c: 79 · 13c: 153 · 14c: 100.
## Columns
- `scholar_id` — the author segment of the work identifier; joins to
`author_uri` in the parent corpus.
- `name_ar`, `name_lat`, `death_year_ah`, `death_year_ce`, `century_ah`.
- `madhhab` — stated for 480 scholars;
empty means the sources do not say, not that he had none.
- `sciences` — space-separated shelves he appears on.
- `n_works`, `n_chars`, `works` — his bibliography as we hold it (`works` is a
JSON-encoded list of titles; parse it with `json.loads`).
- `n_biographical_notices` — how many notices in
[`islamlab/islamic-biographies`](https://huggingface.co/datasets/islamlab/islamic-biographies)
name him (1,344 scholars are
written about somewhere in the corpus). It is a name match, so read it as a
pointer, not a count.
- `source` — which library our copies came through.
## Where this comes from
Derived from
[`islamlab/islamic-sciences`](https://huggingface.co/datasets/islamlab/islamic-sciences),
which is where the gate, the provenance and the credits live: scope is
**Ahl al-Sunnah wa'l-Jamāʿah**, and a work enters only if its author died by
1350 AH, so every text is in the public domain. The machine-readable
transcriptions carry **CC BY-NC-SA 4.0**, and so does this.
Corrections and removal requests are welcome on the **Community** tab.