| --- |
| license: cc-by-sa-4.0 |
| tags: |
| - marvel |
| - wikitext |
| task_categories: |
| - text-generation |
| language: |
| - en |
| size_categories: |
| - 100K<n<1M |
| --- |
| |
| # EDITH Marvel corpus |
|
|
| 202,171 cleaned records - characters, teams, locations, items, events, comics - |
| derived from Marvel Database (marvel.fandom.com) and English Wikipedia, crawled |
| 2026-09-04 via the MediaWiki API. |
|
|
| Used by **https://github.com/Tomionkkas/edith** for two things: training the |
| EDITH model, and building the BM25 retrieval index the terminal answers from. |
| The index is not distributed - it is a pickle, unpickling executes arbitrary |
| code, and it rebuilds from this corpus in about thirty seconds. |
|
|
| ## Shape |
|
|
| Each record is two layers. Schema lines above - `Page:`, `Created by:`, |
| `First appearance:`, `Real name:`, `Codename:`, `Kind:`, `Reality:` - and prose |
| below, under `History:`. The retrieval system reads the first half; the model |
| was trained on both. |
|
|
| Alternate continuities carry their reality in the headline: |
| `Doctor Strange (Earth-199999)` against a plain `Doctor Strange`, which is |
| Earth-616, the unmarked default. |
|
|
| ## Known defects, measured |
|
|
| - ~2.7% headline collisions where the source page states no reality |
| - `;` inside `[[links]]` splits a field (~0.07% of records) |
| - 25 malformed `Reality:` lists |
| - 34 records open on a bare pronoun |
|
|
| ## License |
|
|
| CC BY-SA 4.0, **inherited not chosen**: both sources are CC BY-SA and this |
| corpus stores their prose close to verbatim. Redistribute it, or anything |
| derived from it, under the same terms and with credit to Marvel Database and |
| Wikipedia. |
|
|
| ## Disclaimer |
|
|
| EDITH is an independent project and is not affiliated with, endorsed by, or |
| sponsored by Marvel Entertainment, LLC or The Walt Disney Company. |
|
|
