owned this note   changed 6 years ago  
Linked with GitHub

34c3 Open Science/Research Workshop

Day 2 (Dec 28) 11:00–13:00 in Hall 3 (CCL, 1. Etage)

https://events.ccc.de/congress/2017/wiki/index.php/Session:Open_Science/Research_Workshop

In this workshop about Open Research/Science we want to first identify what we can do to improve any aspect of Open Research/Science (like Open Access, Open Educational Resources, Open Peer Review, Open Data, …). We then pick one or more of these aspects and start working on solutions (Hackathon-style).

The workshop is an excellent place to connect with other people who are interested in making research and science more open.

Bring your own ideas and make sure to add them to this document (even better if you do so before the event)!

Participants

What do you understand under "Open Science"?

Possibilities:

  • Open Education
  • Open Access to publications
  • Open Access to (almost) all data/software used for publications
  • Reproducible Research
  • Wikipedia: "Open research is research conducted in the spirit of free and open-source software. Much like open-source schemes that are built around a source code that is made public, the central theme of open research is to make clear accounts of the methodology freely available via the internet, along with any data or results extracted or derived from them."
  • The "Open" in science should IMHO try to reflect the four principles of openness in software but for publications:
    • ability to use (read/downlad),
    • to study (full access with instructions on the sources),
    • to adapt (build on the research data, use the research data with different analytical methods, translate the publication)
    • to distribute (print, sell, whatever)
  • Open infrastructure available for everyone wanting to do research, in computer science e.g. access to computing reseources, access to ordinary working space, some special web service, etc.)

How to improve Open Science/Research?

We could do some research about

  • public statements from research institutions on what they think "Open Access" is and what it actually is
    • so they are better informed and stop making bad or wrong contracts with private publishing houses
    • can't claim that they do "open access" on their websites
  • gather further notions of "Open access", especially those accepted in all EU member states
  • gather further notions of "Open Science"

With that we could target "außeruniversitäre Forschungseinrichtungen" (those doing programmatic research) and force them (via the Bundestag and the BMBF) to gradually adopt a special notion of "open research" (starting with some of their research projects).

What i found is that, if one really understands the "Berliner Erklärung for Open Access" from 200? - and now finds statements on institutions website that they support that declaration. But sometimes even on an institutes website you find out that this institutions really has no idea what they are talking about. See therefore also, this comment on 10 years after the declaration by the MPI.

Tools:

  • Open Peer Review: Tool that anonymizes author names to eliminate biases (gender, ethnic, origin, …)
  • Open Access: create a dat with all PDFs of CC-BY-licensed articles on CrossRef (better accessibility, easy text and data mining, …)
  • Tool/Compilation/List of all available scientific positions in research related centers

Awareness (might depend on science area):

  • Create a short flyer why to choose CC0 for research data in a language scientists can relate to.
  • Compilation of services / websites which offer open-source, open-data, open-access hosting
    • Supported licences
    • Implications with traditional publishers regards first-publication or exclusive-publication of scientific results or similar
  • Raise awareness and promote a scientific culture of talking or publishing about failed investigations, too. Not only publish successes, as that seems still the common approach in established research institutes. Maybe even create "fail" repo for "fails in methodology", "fails in literature research", "fails in data collection" or "fails in data publishing", etc.
  • Add your idea here

Questions from the internet

Projects

Repository of OA fulltexts

  • Participants: André, Paul, Katrin, Michaela, Michael, Emanuil
  • Goal: Create a repository with fulltexts of Open Access publications.
  • Intermediate Hackathon goal: Create a tool that downloads OA fulltext documents from one publisher (in a way that we can extend it to more publishers).
  • Next meeting: Day 2, 18:00, Freifunk Assembly -> we are currently in CCL level 2 hall 12
  • https://github.com/andrenarchy/libfulltext
  • Python 3!

Draft API

  • Fetch DOIs/metadata from CrossRef:
  • CrossRef:
    • CC-BY set in OAI-PMH API? -> OAI-PMH for subscribers only :(
    • REST API: example request Elsevier CC BY 2016 (used filters might need further testing) https://api.crossref.org/works?facet=published:*,license:creativecommons.org/licenses/by/,container-title:*&filter=member:78,has-license:true,from-pub-date:2016-01-01,until-pub-date:2016-12-31
  • crossref-get-dois -m MEMBER_ID -a AFTER_DATE -b BEFORE_DATE -l CC-BY
    • Target: Scrape all DOIs in one licence and from one publisher
    • Result: One DOI per line on stdout
    • Prefixed with a namespace (doi:, arxiv:, …) 
      ​​​​​​​​doi:10.000/XXXX
​​​​​​​​doi:10.000/YYYY
  • get-fulltext -c CONFIG_FILE -d FULLTEXT_DIR -m METADATA_DIR -f FAILED_FILE
    • Reads IDs from stdin
    • Target: Download metadata, download fulltext (pdf)
    • Plugin-like system for publishers
    • METADATA_DIR/doi:10.000/XXX/crossref.json
    • METADATA_DIR/arxiv:1208.0264/arxiv.xml
    • FULLTEXT_DIR/doi:10.000/YYY/fulltext.pdf
    • FULLTEXT_DIR/arxiv:1208.0264/v1.pdf
    • Later: Supplementary information could be dropped in the fulltext directory as well.
    • Steps:
      1. First check ID prefix (ID Prefix handler)
      2. Download metadata
      3. From first two decide publisher
      4. Determine plugin/module to use for download from publisher
      5. Initiate download
    • Check and implement later:
      • API calls for mass fulltext download? (i.e not one by one, but bulk)

external APIs

Existing useful projects

Paper2HTML

  • Goal: Convert paper .pdf files to .html files as a markup language allows integration of various semantic vocabularies to describe / annotate the contents of a paper.

Paper2Text

Data licensing flyer

  • Participants: Jochen, TODO
  • Goal: TODO

Networking

some places to get information about open sciency things

10.1016/j.physletb.2017.11.066

Select a repo