This panel can also be viewed on YouTube here.
Digital Value, Data Rights, Extractivism & Indigenous Knowledge Issues in Data Mining: An Introduction to the Session
Stephanie von Gavel
CSIRO
Link to references here.
The Significance of Indigenous Data Sovereignty and Indigenous Data Governance
Stephanie Carroll
University of Arizona
Links from slides here.
Extracting Indigenous Resources - Data and Equitable Benefit Sharing
Māui Hudson
University of Waikato
Link to references here.
FAIR's Fair? Local Extraction, Global Value
Neil Davies
Gump South Pacific Research Station
Link to references here.
From Extractive to Generative: Environmental Assessment Data & Implications for Indigenous Data
Paul Box
Commonwealth Scientific and Industrial Research Organisation
Links to references here.
Comments 12
Welcome to the CEDS MiniForum session on Values, Rights, and Culture, where we will explore the social impacts and implications of extracting data from one community and generating value for a different community. Session Moderator, Stephanie von Gavel, introduces featured speakers and provides an overview of how data is valued by different actors within a supply chain context; Stephanie Carroll defines Indigenous data and describes the challenges that Indigenous communities face in securing their data rights and governing their data assets; Maui Hudson discusses the extraction of Indigenous knowledge resources in the context of equitable data sharing; Neil Davies explores genomic research data extraction, digital rules and rights, the economics of knowledge, a tropical island’s digital twin, and how some open science principles widen the global inequality gap; and Paul Box describes the national environment assessment data ecosystem for Australia and implications for Indigenous data and knowledge use.
Please leave your comments and questions, and let us know what data extraction issues YOU THINK the global data research community should prioritize by taking our poll: https://forms.gle/Mjt1CpGkgLyRtUXY9
I really enjoyed all talks in this session. Indigenous data rights is such an important aspect and it is fascinating to think of different ways this is institutionalised internationally but also in individual communities/countries. I would really like to read more about experiences in different contexts, and one aspect I am particularly interested is Indigenous demographic data. Governments conducting censuses, also counting Indigenous population, and population estimates and projections are crucial in planning for services/evaluation of programs etc. But Indigenous communities often can access very aggregated data on a scale of nation/region. I want to explore different ways Indigenous demographic data are constructed, owned, and managed on the ground.
Hi Julia, great points. I suggest the book Indigenous Statistics by Maggie Walter and Chris Anderson, the book Indigenous Data Sovereignty edited by Tahu Kukutai and John Taylor, other pieces by Maggie Walter and Tahu Kukutai. You can find some of these linked from the resources page at GIDA-global.org. Also here in the states Desi Rodriguez-Lonebear does tribal pop projections, she just defended her dissertation at the University of Arizona which includes some of this work. She’s now at UCLA.
I would like to than the organisation to make available such a huge platform and also a great Presentation. I think, the global research community should prioritise privacy protection of an individual. To me it means rest of the issues are equally important for us but privacy protection first.
Kia ora Māui,
Thanks for a really interesting presentation. I think your point about unrestricted openness being an opportunity for corporate extractivism is one that has some striking parallels with open source software and some versions of creative commons (e.g. IBM using CC-licensed photos to train facial recognition ML systems). Are there any examples you draw upon of protected commons as potentially useful alternatives to the ideology of openness?
Kia ora Sy,
There are four general approaches that I’m aware of but am definitely interested in understanding other ways to approach this challenge. First, controlled access databases like the Integrated Data Infrastructure (IDI) in New Zealand but that is to provide access to de-identified government administrative data. Second, in some of the genomic databases where most datasets are open, provision can be made for specific datasets to require permission to access them from the person that lodged them. Third, the use of data trusts to license datasets generated by collectives. Fourth, the use of TK labels to recognise provenance while allowing data to remain in open environments.
na Maui
Hi Sy and Maui ,
within the CC framework there are some useful and, I belive underutilised, licensing options, that are nuances of ‘open’ that can be used to constrain data access based on usage context. These licensing patterns together with several of the other patterns Maui highlighted should, I beleive, be used in combination to address the challenge of enabling the CARE-ful handling of Indigenous data through community driven data governance, more fine grained control over access and use of data, wrapped around data infrastructure. The required appoach lies between the completely open space of public data (commons) and the private (data) goods space, covering administrative/statistical data in government hands, data he;d by commerical entities and unshared personally held data.
The ‘open by default’ and ‘as open as possible’, stance to building data commons, in which I strongly believe, are appropriate for certain kinds of data, and result in the almost exclusive use of CC-BY – anyone can the licensed material for any purpose provided they provide an attribution to author/creator of the resource. The CC-NC – non commercial use licence means that the licensed material cannot (at least under the terms of the CC licence) be used for commercial purposes. .
Sharing data under CC-NC licence does not exclude commercial use, it means that commercial users must seek a different (commercial) licence to use the material (as per the second pattern referred to by Maui) . Adding caveats or extensions to CC-BY and/or CC-SA specifying acceptable usage could also be considered to further permit or constrain use for specific purposes. However increasigly bespoke and varying licences and additional licensing frameworks create insitutional and legal interoperability challenges downstream (as is the case for viral software licesing). A key challenge relates to ensuring compliance with licensing conditions as these approaches are trust based relying on users to do the right thing and non-compliant use is very hard to track with difficulty in challenging misuse when it does occur. However, these issues stem from the nature of digital data rather than the licensing mechanism itself.
A use differentiated licensing approach potentially creates transactions costs for the data holder (in dealing with access permission requests) and the commercial interest seeking to access and use the data and will, I am sure be subject to criticism of it being a drag on innovation and limiting the ability of ‘the market’ to solve challenges. However, from a practical rather ideological perspective, the criticism of increased transaction costs can be mitigated through the use of common data infrastructure for storage of and managing access to collectivised data, together with appropriate data governance arrangements to administer agreed data sharing (Maui’s third pattern) and usage rules and protocols (e.g. CARE and TK principles) Maui’s fourth pattern.
As Maui indicates there are some interesting integrated organisational and institutional forms such as cooperatives and trusts that could be used to manage collective data rights on behalf of the owners/coproducers of the data, acting as intermediary with potential users. Collectivisation of data enables network effects and the creation of value in data while maximising the utility of data, ensuring its use is aligned with (data owner) community values and goals, while at the same time potentially establishing revenue stream (through commercial licensing of data) that could contribute to ongoing costs of information infrastructure maintenance.
I’m slow to join the conversation here, having watched the session last week! I’m very interested in the ways that the Traditional Knowledge labels compare to Creative Commons labels. Whereas CC labels don’t, I think, fundamentally challenge copyright, TK labels feel like they’re an attempt to…pull knowledge out of the Western property system, and back to community ownership (or stewardship)? I’m grateful to have had a chance to learn more about them, and I’m looking forward to incorporating some of the work around TK labels in my teaching as we encourage students to look beyond the binary of strict intellectual property rights vs ‘information wants to be free’. Thank you so much for your work!
Dear Organizers, the DIGITAL VALUE, DATA RIGHTS, EXTRACTIVISM & INDIGENOUS KNOWLEDGE is focusing on new topics in data mining , and digital economy. This is a new concepts in Indian context, where digital services are on expansion and under development stage.
Thanks for this opportunity.
Nice lecture.
I like the session but I hope it have a little complexity next time.
Hi Parihar-
What would you like to discuss in future events?