Grants:IdeaLab/Reform of citation structure for all Wikimedia projects
What is the problem you're trying to solve?
The current infrastructure for managing citations on Wikipedia is causing a range of problems which make it difficult for people, particularly academics, to use and respect Wikipedia.
Multiple reforms are needed to update citation management on Wikimedia sites. The below are some reforms which are already in the works. Each of these are independent projects but it would be nice to have some collaboration between these efforts and to plan some things centrally.
- The Visual Editor development community is requested to provide documentation on their vision for the future of Wikipedia citations. Whatever they are doing may interface with any of the below proposals.
- There are multiple tools which automatically generate citations. The creators of these tools intended to generate consistent citations. However, each tool has unwarranted and unintentional variation in the citations it produces as compared to the output of other tools, because there has never been quality control among them all as a group. All tools which purport to give the same citation as others should be standardized so that they actually do this. This has nothing whatsoever to do with en:WP:CITEVAR and should be completely noncontroversial to anyone who sees the problem.
- Citations should be hosted in Wikidata so that they can be called as interwiki links. This is described in Wikicite. A central reference repository would enable a citation used in one Wikimedia project (like English Wikipedia or Wikimedia Commons) to be reused in another project (like non-English language Wikipedias in translation projects). This would build upon established practices of using en:User:Citation bot to generate local citation templates on each language Wikipedia, which is needlessly repetitive and wastes lots of labor.
- Citation templates which fail to inherently identify in a human readable way the source which they are citing need to be deprecated. A bot solution to this is discussed at Replace "cite pmid" with "cite journal" on English WikiProject Medicine.
- Citations need to have a field flagging whether the content to which they link is behind a paywall, and this field needs maintenance so as to change if the paywall moves. Non-paywalled references need their licensing terms flagged, and there is a proposal for this at en:Wikipedia:WikiProject Open Access/Signalling OA-ness.
- There should be a way for academic journals to be able to feed streams of published articles into relevant WikiProjects so that Wikipedia contributors get access to information. This is proposed at Grants:IdeaLab/Open Access Reader.
- Publishers which donate subscriptions to non-free content need better monitoring structure to check how their content is reused by Wikipedians. In academic publishing this would contribute to determinations of en:impact factor or en:altmetrics. In general, Wikipedia needs better partnership infrastructure for all organizations which share their content on Wikipedia projects, but the The Wikipedia Library specifically assists publishers in getting their academic papers cited on Wikipedia. Good partnerships like this should be developed with citation structure reform. There is a project called en:ImpactStory which already tries to report citation on Wikipedia as beneficial to researchers, and generally there is a trend of researchers seeking Wikipedia collaboration but hitting walls due to Wikipedia not being able to support them in doing this.
- Some organizations have requested advice on making tools for their own websites which could generate citations which could be used on Wikipedia. Tools for partners need to be made so that anyone can make it easy to advertise that they want their publications cited on Wikipedia. One organization which has asked about this is the United States en:National Institutes of Health.
- The Wikipedia community is casual about making recommendations for best practices, but when organizations seek partnerships with Wikipedia, they have the power and resources to make changes far in excess of what a single Wikipedian can do. For this reason, there should be recommended best practices for organizations to make citations so that they can follow the best recommendation and not make an arbitrary choice among options of unconsidered varying quality. When partner organizations make choices while not knowing Wikipedia community culture, they can suddenly make large sweeping changes which may not be best for the community. Practically all partner organizations will have some opinion on citations. User:Bluerasberry has some background on this at The Manual of Style for Medicine-related articles, "Full citations versus citation templates".
- There is some talk of actually publishing original research in a Wikimedia project and putting it through peer review as is typical with academic research. The main precedent for original reporting is Wikinews. At v:Wikiversity:Peer review and v:Wikiversity Journal of Medicine there is a proposal and pilot to publish and review original research in Wikiversity. External to Wikipedia, en:Scholarpedia is a project on the MediaWiki platform which publishes and reviews original research as a wiki-based academic journal, and they have established a precedent and model for doing this on Wikipedia. Among other innovations, Scholarpedia assigns en:Digital object identifiers to the articles it publishes.
- The WikiScholar proposal seeks to manage a bibliography and range of metadata and interconnections between sources cited.
- It has been a goal and dream of many websites and every academic publisher to make the social media forum for academic discussion around individual academic papers. If a Wikimedia project hosted a large amount of metadata about every citation, then either in a Wikimedia project or with some kind of port for external website, it would be easy for anyone anywhere on the Internet to tag their comment in any publishing forum so that it is noted as a comment on a given academic paper. Wikimedia projects have traditionally not hosted these kinds of discussions, and this commentary system could be associated with the reliable sources filters of the sort used by Wikipedia, but the existence of centralized citation management on a Wikimedia project is an ideal base for centralizing and cataloging all discussion in any context about any given academic publication anywhere. This would be inspired by en:Web of Science and similar tools.
- More community discussion is needed. A talk happened at Wikimania 2014. In 2015 wm2015:Submissions/Free as in Free: Strategies for Advancing Open Access on Wikipedia addressed some of the same issues.
- @Pigsonthewing: has this summary of what the future will probably be.
What is your solution?
The proposed solution is to break this project into subprojects in such a way that anyone can work on any subproject without too much central coordination. The biggest problem in all of this is getting advice from the Wikimedia Foundation and developers of Wikidata about the future of storing citations on Wikidata, as using Wikidata for this is probably the only reasonable long term solution as tens of thousands of citations - for example to medical journals for medical articles - need to be reused on multiple articles in multiple languages of Wikipedia. This will remain a huge mess for so long as there is not central citation management, for reasons including inability to fix broken citations, difficulty modifying Wikipedia's citation template code into localize for multiple languages, inability to quickly find and replace citations when scientific papers are updated (especially true for medicine), and inability to report citation metrics to journals and academia to increase the value of Wikipedia to established educational channels.
The goal of the project is to make the citation infrastructure on Wikipedia unsurprising. Ultimately this would lay a foundation for much deeper community partnerships with all publishers of academic content, including journals, schools, governments, and all of the standard bearers of the best information.
Welcome, brainstormers! Your feedback on this idea is welcome. Please click the "discussion" link at the top of the page to start the conversation and share your thoughts.