7 min read

Barcamp Open Science

2018/03/12

Introduction

The Barcamp ‘Open Science’ took place on 12 March 2018 in the rooms of the Wikimedia Deutschland, Tempelhofer Ufer 23-24, 10963 Berlin. The event is organised by the Leibnitz Research Allience Science 2.0.

Central aim of the Barcamp is to build a network between people working in the field of Open Science. Opposed to a Conference, the Barcamp concept follows a bottom-up approach.

Among the participants were researchers, software developers, librarians, e.g. from

In the introduction round they named e.g. the following topics:

  • Metadata
  • Open Knowledge Maps
  • Open Methods
  • Open Research Data Quality
  • Research Data Management
  • Training Open Science
  • Transparency
  • Workflows

The Barcamp started with a session planning where topics for sessions of 45 minutes duration were proposed.

Content

Ignition Talk by Lambert Heller

  • Lambert Heller

  • Elsevier suggestion add geoblocking to open access

  • Lessons-learnt: platforms require trust, but often exploit it!

  • 2001: BitTorrent. It turned the Client-Server approach of the internet upside-down

  • 2013: Decentralised web movement (DAT). The “Dat Project” (A distributed data community) started on github.

  • We have a “trusted platforms” problem in science

  • Multiple tools are needed

  • The GHTorrentProject, Debian CD images with BitTorrent

  • Why move schoolary publishing to peer-to-peer (P2P) networks?

    1. In order to get their research done, researchers should be able to get hold of lots of data without additional effort. It is up to the research infrastructure.

    2. We owe it to the scholarly work incorporated in that data

    3. Replacing privileged access with permissionless innovation levels the playing field for business model innovations.

  • Pragmatic school, https:/doi.org/ck99

  • Economic school

Internet Resources

Session: What kind of tools should we use?

https://etherpad.wikimedia.org/p/workshop_OpenScienceFellows_BarcampSession5

Background

Sind Tools wie Hypothes.is und GitHub zu empfehlen? - Are tools like Hypothes.is and GitHub recommended for open science?

https://nextcloud.gbv.de/nextcloud/index.php/s/Rfg199D4xg0RptB

Kriterien

1 Quelloffen (source code is open)

2 Migration muss möglich sein (Exit-Strategie - eingebauter Knopf) - migration needs to be possible

3 Langfristig gesicherter Anbieter ( >10 Jahre) - offering should be long term (more than 1o years)

4 Vertrauenswürdiger Anbieter (Datenschutz, Non-profit, pro-Europe…) - trustful offering (e.g. data security, non-profit, pro-europe…)

5 Wissenschaftlich verlässlicher Anbieter. z.B. Fighshares hat keine Tombstone: https://doi.org/10.6084/M9.FIGSHARE.1381402 - Scientifically reliable provider.

6 Sichtbarkeit (leicht in Google findbar) - visibility

7 Offene Beteiligungsmöglichkeit (externe Nutzer/Accounts leicht möglich) - open participation should be possible (external users/accounts)

8 Usability

A reference that should be considered: Bilder G, Lin J, Neylon C (2015) Principles for Open Scholarly Infrastructure-v1, retrieved [date], http://dx.doi.org/10.6084/m9.figshare.1314859“ - https://cameronneylon.net/blog/principles-for-open-scholarly-infrastructures/

Notes

Reference: https://figshare.com/articles/NPOS_Workflow-perspective-Bosman-Kramer_pptx/5065534/1

Applied Criteria

ResearchGate: 6, 7, 8

Open Science Framework: 1, 2, 5, 6, 7, 8

FigShare 6, 7, 8

Zenodo 1, 2, 3, 4, 5, 6, 7, 8

Overleaf 2, 6, 7, 8

GitHub 2, 6, 7, 8

ScienceOpen

Criteria need to be applied:

arXiv 1?, 2?, 3, 4?, 5?, 6?, 7?, 8?

bioRxiv 1?, 2?, 3?, 4?, 5?, 6?, 7?, 8?

Jupyter 1?, 2?, 3?, 4?, 5?, 6?, 7?, 8?

Authorea 1?, 2?, 3?, 4?, 5?, 6?, 7?, 8?

MyExperiment 1?, 2?, 3?, 4?, 5?, 6?, 7?, 8?

protocols.io 1?, 2?, 3?, 4?, 5?, 6?, 7?, 8?

OpenNotebookScience 1?, 2?, 3?, 4?, 5?, 6?, 7?, 8?

GitLab (CE at institutions) 1?, 2?, 3?, 4?, 5?, 6?, 7?, 8?

Dryad 1?, 2?, 3?, 4?, 5?, 6?, 7?, 8?

Dataverse 1?, 2?, 3?, 4?, 5?, 6?, 7?, 8?

AsPredicted 1?, 2?, 3?, 4?, 5?, 6?, 7?, 8?

Hypothes.is 1?, 2?, 3?, 4?, 5?, 6?, 7?, 8?

Zotero 1, 2, 3?, 4?, 5?, 6?, 7?, 8?

RIO 1?, 2?, 3?, 4?, 5?, 6?, 7?, 8?

Eigene Blogwebseiten 1?, 2?, 3?, 4?, 5?, 6?, 7?, 8?

Twitter 1?, 2?, 3?, 4?, 5?, 6?, 7?, 8

Session: How do we motivate / reward doing Open Science?

https://etherpad.wikimedia.org/p/oscibar2018_session5

Point of View: How open science helps researchers succeed https://doi.org/10.7554/eLife.16800.001

Felix Schönbrodt: https://twitter.com/nicebread303/status/973138967091654656

Discussion https://twitter.com/BrianNosek/status/949756218817630208

Data stories:

Session: Valid reasons for opting out of open science

https://etherpad.wikimedia.org/p/oscibar2018_session13

Background:

Opt-out policy in H2020: ‘projects can “opt-out” of these provisions before or after the signature of the grant agreement (thereby freeing themselves from the associated obligations) on the following grounds:’

  1. Incompatibility with the Horizon 2020 obligation to protect results that are expected to be commercially or industrially exploited

  2. Incompatibility with the need for confidentiality in connection with security issues

  3. Incompatibility with rules on protecting personal data

  4. Incompatibility with the project’s main aim

  5. If the project will not generate / collect any research data, or

  6. If there are other legitimate reasons not to provide open access to research data

Problem:

To general. Too simple to out-out!

Make the reasons more detailed!

Session: CC0/PD vs. (CC)Licences for Open Science

https://etherpad.wikimedia.org/p/oscibar2018_session17

Licenses

Licenses

Source

Session: Pain Points in Open Science

https://etherpad.wikimedia.org/p/oscibar2018_session3

Pain Points:

  • Replication studies are rejected for publication. Impact factor is old style but it is the currency you are paid
  • Often, Open Data are not allowed or researchers do not want to open their data
  • Tools, e.g. git is a pain to use
  • Efforts: it takes money and resources to change things
  • Project proposals are not focused on data

Solution:

  • Allocate money
  • You need supporting structures
  • You need an Open Data Strategy! Strategy could be: publish metadata of sensitive data, together with a contact information. Then, people interested in the data can contact the owner of the data and a contract about what is allowed with the data can be made.
  • Birte Pfeiffer (Research data manager at GIGA, Hamburg): “You have to talk to the researchers! The question must be: How can I help you?”
  • Take care of the data from the beginning -> Data management plan -> is a “living” document

Session: Research software publishing/repositories

https://etherpad.wikimedia.org/p/oscibar2018_session6

  • Welche Probleme gibt es, Software zu finden?
  • GitHub + DOI
  • Allianzinitiative: FAIR principles
  • Eigenentwicklung (wer?): Metadatenportal, nur aggregierte Daten, zugänglich ist ein (noch privates) GitHub repo, Software auf Django basierend
  • Anja Busch, Leibniz-Informationszentrum Wirtschaft: Metadaten-Harvesting
  • Birte Pfeiffer, GIGA Hamburg: Suche nach Datenerhebungs-Tools -> Wie soll ich das alles testen?
  • Sandbox-Anwendungen und Ansprechpartner sind wichtig zum Evaluieren von Software -> using mybinder?

Resources

openscience.org, The OpenScience Project de-rse.org, Research Software Engineers

Session: Metadata / Codebooks

https://etherpad.wikimedia.org/p/oscibar2018_session9

Session: Open Knowledge Maps

https://etherpad.wikimedia.org/p/oscibar2018_session14

Session: Doing research outside academia / Citizen Science

https://etherpad.wikimedia.org/p/oscibar2018_session20

OK-Lab Berlin: Presentation

  • What is Open Data?
    • machine readable
    • open licences
    • without payment
    • sharing and remixing allowed
    • allowed for commercial purposes
  • metadata: additional pdf or something
  • Formats for open data in increasing quality: pdf -> xls -> csv -> rdf -> lod

Data portals

Example applications

References

Personal Contacts

Alexander Struck, HU Berlin: We may have a look at Seafile as a software tool for storing our raw data.

Wrap-up