UK Reproducibility Network
  • About
  • Community
    • Steering Group
    • Local Network Leads
    • Institutional Leads
    • External Stakeholders
    • Advisory Board
    • UKRN Academy
    • International
  • Resources
    • Terms of Reference
    • Code of Conduct
    • Common Statements
    • Speaker Directory
    • Primers
    • Open Research across Disciplines
    • Presentations
    • Materials
  • Activities
    • Events
      • Open Research Calendar
    • Open Research Programme
      • Partner Institutions
      • Training
      • Evaluation
      • Sharing
      • Reward and recognition
    • Initiatives
    • News
  • Join UKRN
  • Contact
Select Page

Contents

  • 1 Open Research across Disciplines
  • 2 Artificial Intelligence
    • 2.1 Case Studies
    • 2.2 Examples of open research practices
  • 3 Resources
      • 3.0.1 General Resources
      • 3.0.2 Open Methods
      • 3.0.3 Open Data
      • 3.0.4 Open Outputs

Open Research across Disciplines

How the principles of open research can be applied to your discipline

Artificial Intelligence

To suggest improvements or additions to this page, please use this form.

Case Studies

UKRN case study: Machine learning, computer vision, and explainable AI

University of Surrey case study: Open research principles for depression diagnosis with machine learning | University of Surrey

University of Sheffield case study: Haiping Liu on PyKale – Accessible machine learning from multiple sources for interdisciplinary research. Video

Examples of open research practices

Open Methods: “Microsoft Research [have] deployed software agents trained with natural language understanding capabilities to continuously scavenge the Web for research artifacts and, from them, extract up-to-date academic knowledge into a graph-based representation called Microsoft Academic Graph (MAG) (Sinha et al., 2015). As the records are from the entire web, MAG equalizes the discoverability of research materials made accessible by the incumbent publishers as well as by individual authors self-archiving at their own websites, potentially making policy initiatives to favour “Gold” over “Green” OA (e.g., Gibbs, 2013) less critical.” (https://doi.org/10.3389/fdata.2019.00026)

 

Resources

General Resources

Implications of openness in AI. https://www.nickbostrom.com/papers/openness.pdf

Open Methods

  • Presentation on new tools and services in AI and open research. https://www.stm-assoc.org/2018_12_05_STM_Week_Innovations_Jon_White.pdf
  • Recommendations on how to create reproducible AI. https://www.isi.edu/~gil/papers/gundersen-etal-aimagazine18.pdf
  • Open data, tasks, and analysis scripts for machine learning.
    • https://www.openml.org/
    • https://arxiv.org/pdf/1402.6013.pdf
  • Open-source software and tools. https://openai.com/

Open Data

Comprehensive list of machine learning datasets. https://lionbridge.ai/datasets/the-50-best-free-datasets-for-machine-learning/

Open Outputs

Preprint repository. https://arxiv.org/

 

This page is adapted and extended from: Farran, E. K., Silverstein, P., Ameen, A. A., Misheva, I., & Gilmore, C. (2020, December 15). Open Research: Examples of good practice, and resources across disciplines. https://doi.org/10.31219/osf.io/3r8hb

About UKRN

UKRN is a peer-led consortium that aims to ensure the UK retains its place as a centre for world-leading research. It is led by Marcus Munafò (Bristol), Chris Chambers (Cardiff), Alexandra Collins (Imperial), Laura Fortunato (Oxford), Etienne Roesch (Reading), and Malcolm Macleod (Edinburgh).

Recent posts

  • Dorothy Bishop Prize 2023 Winners
  • Indicators of Open Research: UKRN call for priorities
  • Open Research Programme welcomes new Institutional Partner

Subscribe by email

  • Twitter
© UK Reproducibility Network. All rights reserved.