3rd-BEXIS-2-User-and-Developer-Conference.md

About・Abstract・Dedication・Contact

Opening up the research data life cycle

Abstract

More and more aspects of the research data life cycle involve some form of sharing - be it the protocols used to generate the data, the software to analyze it, metadata about the data, the data itself or other aspects of its life cycle, from ethical review to citizen science. This talk zooms in on some examples in different stages of different data life cycles and explores how the changing dynamics of data sharing interact with the dynamics of doing research more generally.

The research data life cycle

• can be framed in many ways
  • (UVA example)
• key components: plan ・ get permissions ・ get funding ・ get infrastructure ・ collect metadata ・ collect data ・ process data ・ analyze data ・ publish ・ curate data ・ preserve data ・ reuse metadata ・ find data ・ access data ・ reuse data ・ reproduce data ・ delete data
• exact arrangement can and does vary
• frequent 'rinse and repeat'

What does it mean to open it up?

• no single publishing step within the cycle, since every intermediate step is an opportunity to share
• video

Some of my activities in this space

• helping to improve data sharing in public health emergencies
• working on making data mangement plans machine actionable, versioned and public
• helped shape the Open Science Prize
• helping to explore technical aspects of preprint services in the life sciences
• helping to collect citations for the sum of human knowledge
• contributing to data science policy in the European Union
• helping to organize events around modernizing scholarly communication
• editing an open-science journal that aims to make entire research processes visible, not just the results
• running a bot that feeds openly licensed multimedia from PubMed Central into Wikimedia Commons for reuse on Wikipedia and elsewhere
• editing the Topic Pages collection at PLOS Computational Biology

How to discover research that is being performed now or planned for the near future?

• Actual need
  • Zika example
  • further notes on emergency response
• Some inspiration
  • real-time visualization and sonification of Wikimedia edits
  • What about doing this for research?
    • As it happens (e.g. as per the video above)
    • All along the research cycle?
  • with appropriate filtering, something of this kind could serve as the basis for researchers, funders, the public and others to engage with a given research topic
  • particularly relevant in the case of public health emergencies, which also provide a context where "open by default" is actually within reach
  • works best if data and metadata are FAIR by default (see open consultation)
  • publicly versioned machine actionable data (or project) management plans as a way to connect research projects (past, ongoing and future) with relevant individuals, communities, institutions and their respective resources and workflows
• Queriability: How can data from different parts of the research ecosystems best be integrated in a way that allows to address issues that cut across research fields?
  • Wikidata as a hub for structured open knowledge across domains, with biomedicine being one of the pilot areas
    • sample queries
      • map of topics co-occurring with "Zika virus" in the literature
      • scholarly profile pages
• Reproducibility
  • Jupyter notebook from LIGO
  • analysis of reproducibility of Jupyter notebooks
  • today's workshop Fostering reproducible science – What data management tools can do and should do for you
• Ethics: making ethics data FAIR
  • talk on the topic
  • further notes
• Sustainability: How to better align the priorities within the research ecosystems (open or not) with the needs of humanity as a whole?
  • e.g. the UN's Sustainable Development Goals
  • further notes
• Lean administration: How can openness be used to reduce bureaucracy?
  • e.g. self-organized fund allocation
• Effective policies: How can different policies and infrastructures be made better aware of each other?
  • make policies machine readable?
  • start with machine readable licensing statements, e.g. CC0 waiver
  • dedicated talk

What does this mean for BEXIS?

• Use persistent identifiers whenever possible
  • botanical collections example
• Use open licenses and avoid -NC restrictions to the extent possible, and CC0 for data
  • WikiPathways example
• Share early and often
  • let others see more of what you are doing, closer to when you are doing it upcoming iDiv symposium on data mobilization
• Share where others are looking to find things
  • Ernst Abbe example
• Others includes people and machines
• Your ideas

Dedication

I dedicate this talk to the memory of my grandfather Werner Mietchen, whose funeral is taking place the day after.

Contact

• @EvoMRI
• email: daniel.mietchen@virginia.edu
• ORCID: 0000-0001-9488-1870

About

This page ( https://github.com/Daniel-Mietchen/events/blob/master/3rd-BEXIS-2-User-and-Developer-Conference.md ) hosts a presentation that was given on 15 June 2017 at 7am UTC. It is a contribution to the 3rd BEXIS 2 User and Developer Conference taking place on 16 June 2017 in Jena, with an associated reproducibility workshop on June 15 in conjunction with Data Science Day 2017. BEXIS (also BExIS) stands for "Biodiversity Exploratories Information System" and is a research environment for managing the entire data life cycle for biodiversity data.

Some impressions from the event are available here.