Ideas for what to dump:
- enwiki-20181107-ores-scores.jsonlines.bz2 – Full dump of all ORES scores available for the wiki.
- enwiki-20181107-ores-scores-editquality.jsonlines.bz2 - Dump of all scores for a specific model.
- wikidata-20181107-ores-scores-itemquality-history-monthly.jsonlines.bz2 - Dump of all historical scores for a content quality model, but sampled down to one revision per page per month.
- wikidata-20181107-ores-scores-itemquality-current.jsonlines.bz2 - Much smaller dump that only has scores for the latest revision of each page.
These will require some processing to create, a MapReduce query should be fine.
The last time we made a dump like this, it was created manually. About 300 people have downloaded and it's cited by 2 papers, so it seems to be useful.
https://figshare.com/articles/Monthly_Wikipedia_article_quality_predictions/3859800