Timeframe
2019–2024
Funding
Ministry of Business, Innovation, and Employment
COMPASS staff
Barry Milne
Eillen Li
Collaborators
Victoria University of Wellington
Colin Simpson
Binh Nguyen
Tom Elliott
Robin Blythe
Alex Wang
iNZight Analytics
Andrew Sporle
Description
This multi-institution data science research programme led by Professor Colin Simpson at Victoria University of Wellington focuses on informatics for social services and wellbeing. The primary aims are to:
- Improve data standards;
- Promote Māori data sovereignty;
- Develop systems to support access;
- Evaluate synthesising of datasets;
- Identify security and privacy implications; and
- Investigate the use of machine learning and AI methods.
Full details of the programme can be found at https://terourou.org. It addresses the critical needs for better access to datasets and better linking of data in New Zealand, leveraging some exceptional administrative data resources that were already out there, but that people did not know so much about.
We are looking to:
- Enhance current data platforms providing nationwide access to a range of large, in-depth datasets with information on social services;
- Create an enhanced inventory and description of the different data sources available in New Zealand and associated metadata;
- Develop solutions that encourage data privacy and ethics, data standards, and Māori data sovereignty
- Clarify legal issues, and develop and apply novel data science and analysis approaches.
The research programme will help in the production of better policy, and enable enhancements in the delivery of social services in a range of areas. It will also increase the number of research projects using New Zealand’s rich social data infrastructure, and their diversity, and add value in these areas.
We will increase the profile of social data and highlighted its importance and research impact and, with a focus on best practice, help to ensure the best chances of buy-in from a range of academic and government stakeholders. COMPASS staff have worked specifically on:
- describing linkage bias in the construction of the IDI
- what is possible with regard to the analysis of intergenerational data in the IDI; and
- solutions to data access by way of creating synthetic datasets.