General Resources
Please find resources below to assist in preparing a competitive proposal and more generally to support the creation of quality training and evaluation datasets. General resources are below. For domain-specific resources, select the Agriculture, Language, Health, or Climate resource tabs
Lacuna Fund Policies
Lacuna Fund Guidance
- Dataset Hosting and Documentation Guidance
- Guidance on Environment, Conflict, and Gender
- ML Dataset Sustainability Toolkit (prepared by LDRI)
- Community Engagement Framework for Machine Learning Datasets
Lessons Learned
- 2023 Lacuna Fund Learning and Evaluation Report: South African-based consultants Data Innovators performed an evaluation of Lacuna Fund’s activities and funded projects from 2020-2022. This report assesses what aspects of Lacuna Fund have effectively and efficiently enabled the creation, expansion, and maintenance of representative and unbiased training datasets for ML; examines process challenges experienced by stakeholders; and provides recommendations for improvement.
- INFOGRAPHIC – 2023 Lacuna Fund Learning and Evaluation Report: Summary of measurable results, lessons, and best practices from implementing Lacuna Fund over 2020-2022, providing insights on achievements and lessons for the ongoing improvement of the fund.
- 2020 Grantee Lessons Learned: Summary of lessons learned from the first round of Lacuna funding, which included the Agriculture and Natural Language Processing domains.
- Lessons Learned from Lacuna Fund’s First Year of Funding: This report summarizes outcomes and learnings from Lacuna Fund’s first two calls for proposals with an eye towards providing guidance to future applicants and insight to the field of machine learning and artificial intelligence for social good.
Lacuna Fund in the Press
Cloud Storage and Computing Power
- In-kind cloud storage and computing power may be available from Lacuna Fund partners. If you would like to utilize this resource, please include it in your budget. Selected teams will receive instructions for how to apply when they receive their award. If you are interested in including in-kind services in your proposal and budget, the following tool is available to help estimate the cost for your project: https://cloud.google.com/products/calculator
- Machine Learning CO2 Emissions Calculator: https://mlco2.github.io/impact/#compute
Equity
- Principles for Using Public Health Data to Drive Equity (CDC Foundation)
- Gender equality and inclusion (GEI) glossary (IDRC): Presents concise definitions for core concepts central to IDRC’s efforts to advance GEI through research. Also available in French | Spanish.
Privacy
- Anonymization tool (Google): De-identifying sensitive data | Data Loss Prevention Documentation | Google Cloud
- UK Privacy Enhancing Technologies (PETs) offers ways to strengthen privacy measures within datasets to make the data more shareable: PETs Adoption Guide (cdeiuk.github.io)
- Data Privacy and Access: Overcoming Barriers to Data Sharing (PPT)
- Data Anonymization Guide (Lacuna Fund)
- A Visual Guide to Practical Data Deidentification
Satellite Imagery
- Moving from awareness to action with the Nonprofit Program (planet.com). Contact Planet at go.planet.com/nonprofit for more information.
Presentations & Webinars
- Toward Equitable Datasets and Data Sharing for AI in Agriculture Webinar (15 October 2020)