Annotated Hub of Science

From Meta, a Wikimedia project coordination wiki

The Annotated Hub of Science is a proposal for an integrated stack of documents, data, summaries and discussions, and social networks related to science. Science is only as good as its tools and methods; including its methods of discovery, collaboration, refinement, and delight.

Please edit, spindle, and refactor at will.

Stack[edit]

  1. Direct access to research
    1. "Hillel", a P2P repository: An app which lets authors host (locally, on their phone/computer) their published work, reimplementing researchgate in a robust + (c)-compliant way, whether or not they have an institutional repository. (via IPFS.)
    2. "Starling", an #ICanHazPDF service: A mechanism for people to request an article & have the request added to a global list. These requests can be processed by any script or person (like 3.), can update collective bounties, &c. Starling gets you a copy eventually, even if the article isn't yet OA. In some jurisdictions requests can be amplified by a FOIA request for the article, when one author is at a public university : in this case, as with Hillel, a software agent in the client can catch and respond to recurring requests, so the author only has to find and post an article once.
  2. Overlays, summaries: "Distillation", "I can't believe it's not Science/Nature!" --
    1. "Distillation" -- dual documents, summaries that stay close to the original to work w/ annotations of them in-place
    2. "ICBINSN" -- A universe of dual journals, w/ the same list of articles, echoing {both closed + open} journals, offering automated and streamlined summaries + commentaries.
    3. These can contribute to Starling.
  3. Metrics: "Public Web of Science" - free + customizable metrics for published literature: extending the good work of Scholia and PLOS.
    1. Use a focused Wikicite wikibase w/ 1 entry per {citable work [articles, patents, books], authors, topics, affiliations, funders}.
    2. Capture curation work from institutions. [e.g.: Dimensions gets data donations from libraries + funders]
  4. Annotated reader: "Sci co-analysis"? -- A reader with an annotation overlay for reading papers
    1. Helps you search for a paper [via UPW/Starling] or reads one you give it, overlays a reasonable auto-curation of public notes (Tweets, blog posts) on the article, lets you add your own. (many related prototypes exist -- Codex may be the best)

Related efforts[edit]

Things we probably don't need to do ourselves:

Some of these we could use as our backend/source or overlays.

Components[edit]

Storage
  • DocumentCloud can store O(100M) documents + integrate w/ sunshine requests.
  • Wikidata (or a dedicated Wikicite instance of wikibase) can store associated metadata + sources
  • Zenodo + a Dataverse instance can store arbitrary datasets [like MAG] until they can be parsed into the above. DV supports shared folders for projects.
  • Hypothesis? can store the annotations themselves
Interfaces
  • For discovery: [ICBINS, ...]
  • For sharing one's work: [FMe, ...]
  • For opening other work: Dissemin, OAButton,
  • For annotating:
  • For comparing one's school/dept: [dissemin, OpenAire dashboards, ]
  • ... wanted
Feeds
  • The Crossref Event Feed can incorporate and share feeds of new publications
  • Artefacts models of capturing drafts, nanopubs, methods, granular steps. ...
  • ...
Support
  • Technical: transfer, digitization, amberizing
  • Editorial: flipping (catalog, field contacts)
  • Legal: process, institution, regional
  • Admin: consortia, nets
  • Publicity: docs, stories