Wikicite/e-scholarship/Evolution and Evolvability (WikidataR for Wikicite)
Project name
[edit]WikidataR: Expanding read-write capabilities for R to write enhanced publication metadata to Wikidata
Contact information
[edit]- application type
- your name and/or Wikimedia username
Evolution and Evolvability (Thomas Shafee) - Melbourne
Thomas.Shafee@gmail.com
- Your nearest city and country
Melbourne, Australia
- Details of team members (optional)
If this application is for a team project, each additional team member (up to 5 total) should have their name/username, project role, location, and contact details here:
- Canley (Alex Lum) – Melbourne: co-developer, experienced in R programming, SPARQL and QuickStatements. Email: alex.lum@wikimedia.org.au
The activity/project
[edit]Tell us about your proposed project. What will you get achieved with this time?
Current situation
The scripting language R is the most used data science language, well suited to. There has been a R packages around for a while to read from wikidata (reading single items with WikidataR and submitting SPARQL queries with WikidataQueryServiceR). However no package exists that can write to wikidata from R.
Proposed action
I am therefore creating an expanded and updated WikidataR to have this ability to write to Wikidata via the quickstatements API (by expanding the functionality of the old read-only WikidataR package, since its original creator is no longer a developer).
Applications
The ability to write to Wikidata from R will greatly enhance the community's ability to enrich publication items in Wikidata. These include properties such as funding scheme (P6195), editor (P98), reviewed by (P4032), or describes a project that uses (P4510) for methods and equipment; significant events such as ethical approval (Q98550700) or rectification (Q56478729); and author/editor/reviewer properties such as declared conflict of interest (Q99429881). Similarly, journals can be indicated as predatory publisher (Q65770389) or hijacked journal (Q17300027). As a language, R is particularly well suited to the sort of data handling necessary to cross-reference existing databases and write to appropriate wikidata items, and this R package will act as an equivalent for the to the python PyPl package.
Your qualifications
[edit]Describe how you (and, if applicable, your team)are able to achieve this project. What skills, expertise, and motivation do you have which will enable you to succeed?
- I have experience in publication metadata open metadata from my work with the WikiJournal User Group
- I have experience in research output metadata (including non-traditional research outputs)
- I have experience in R package development and basic web interfaces
- I have already implemented some basic proof of concept functions and tested them:
- read functions to retrieve QIDs from DOIs and ORCiDs.
- write functions to edit existing wikidata items (though not yet create)
Proposed activity dates
[edit]Four days total:
- 10-11 Oct 2020
- 17-18 Oct 2020
Budget (Standard WMF per-day living allowance - 75% of the "M & IE rate" for relevant city)
- US$145 x 4 days x 2 people = US$1160 x 0.75 = USD $870 ( ~$1220 AUD)
Endorsements
[edit]Optional: Community members are encouraged to endorse your proposal and leave a rationale here.
Support R is big, so a two way street would be good value. Proposer is also a valued and valuable member of the community with a strong history of good contributions. --99of9 (talk) 12:15, 24 September 2020 (UTC)- Support A useful development with potential to increase contributions and contributor base Pru.mitchell (talk) 23:38, 24 September 2020 (UTC)
- Support The R community includes many scholars that could add useful contributions to wikidata. An R package that can write to the wikidata endpoints enables us to create automatic tools in a language many people are proficient at. I also think this has the potential to increase contributions and contributor base. RMHogervorst 2020-09-25
- Support This could provide great exposure for wikicite to the large R community - and would produce a useful utility. Jodi.a.schneider (talk) 20:11, 25 September 2020 (UTC)
- Support R is very important data science language, and R package that can write to the Wikidata it'll add a great value for the project. And as said above, that Thomas is a valued and valuable member of the community. Best --Alaa :)..! 12:36, 27 September 2020 (UTC)
- Support R package can provide useful tools and applications. Yes for sure.--Michel Bakni (talk) 12:42, 27 September 2020 (UTC)
- Support Nawaraj Ghimire (talk) 02:57, 28 September 2020 (UTC)
- Support. Interesting package worth developing. --Csisc (talk) 22:01, 30 September 2020 (UTC)
...
- Status
- complete