data flow

Data Sources

1. PMC
2. Crossref
3. Undetermined sources:
   1. OA button
   2. OCLC
   3. ImpactStory
4. Wikipedia:
   1. Most cited journals on enwp
   2. Most cited articles on enwp

Wikimedia Projects

1. Wikimedia Commons
2. Wikisources
   1. Primarily English language.
   2. Identify the 1000 most cited articles on enwp, import the suitably licensed ones into enws
3. WikiData
4. Wikipedias
   1. Especially English language.
   2. Eventually, all ;)
5. WikiTech (for bots/tools)
   1. Including Wikimedia Labs (virtual private servers).
   2. Including Wikimedia Tool Labs

Phase 1

1. Import media files from PMC to Wikimedia Commons
   1. Will eventually trigger a record in Wikidata, but not now.
   2. Add relevant categories to media files.
2. Import full text content from PMC to Wikisource (or equivalent)
   1. Wikisource item submission will automatically (in a delay) create a record in Wikidata
   2. Embed article media from Wikimedia Commons.
3. Import article license data from CrossRef to Wikidata.
   1. Consider verifying CrossRef license data against PMC license data (flag inconsistencies).
4. Bots run from Wikimedia Labs servers on English Wikipedia to:
   1. Implement Cite DOI templates
   2. Signal license in template
   3. Signal content in template from:
      1. Wikisource
      2. Wikimedia Commons
      3. Wikidata
   4. Link to third party services in template to:
      1. OA Button
      2. ImpactStory
      3. PeerLibrary
      4. OCLC

Phase 2

1. When importing media files to Wikimedia Commons, check consistency and formatting of article meta data from CrossRef.
2. Check metadata between Wikimedia Commons files and WikiData records.
3. Expand bots to run on all language Wikipedias.
4. Include more experimental components to the Wikipedia cite template, such as dependent on services run on Wikimedia Labs servers:
   1. E.g. alt-metrics