This thread will operate as a place to document progress in creating AuthorBot: A bot that automatically adds and corrects author names (and eventually author pages?) to scientific articles on Wikidata.

To see data regarding the most popular scientific databases on Wikidata, go here:
https://github.com/feliciss/wasian/blob/main/wasian/wikidata-database-analysis/graphs/
Some graphs of note:

Progress on AuthorBot:

• Bot queries scientific articles with missing author information:
  • Articles where P50 items are missing a P1932, P9687 or P9688 qualifier
  • Articles where P2093 items are missing a P9687 or P688 qualifier
  • Articles with no P2093 and P50 items
  • TO ATTEMPT LATER:
    • Articles with only P2093 items that could have author name strings replaced with authors
    • Articles with P2093 items that have author initials that could be replaced by full author names (Articles with a PubMed ID good candidates for this)

• Bot is able to obtain and parse citations from a variety of academic databases:
  • PubMed ID (P698)
  • PubMed Central ID (P932)
  • Dimensions Publication ID (P6179)
  • CJFD Journal ID (P6769) - Very difficult. Attempt later.
  • ResearchGate ID (P5875)
  • ADS Bibcode (P819)
  • Bonus Databases:
    • DBLP Publication ID (P8978)
    • arXiv ID
    • OpenCitations Bibliographic Resource ID (P3181)
    • JSTOR Article ID (P888)

• Bot is able to automatically add author names to articles from a variety of academic databases
  • PubMed ID (P698)
  • PubMed Central ID (P932)
  • Dimensions Publication ID (P6179)
  • CJFD Journal ID (P6769)
  • ResearchGate ID (P5875)
  • ADS Bibcode (P819)
  • Bonus Databases:
    • DBLP Publication ID (P8978)
    • arXiv ID
    • OpenCitations Bibliographic Resource ID (P3181)
    • JSTOR Article ID (P888)

• Bot is able to automatically detect if an author page exists on Wikidata and add it instead of an author name string to scientific article pages on Wikidata

• Bot is able to automatically create author name pages from author name strings

Some thoughts:

• Both PubMed IDs and CJFD IDs should be prioritised due to the sheer volume of articles with said IDs and capacity to parse Chinese names accurately, respectively.
• CFJD IDs are so horrible to deal with!