Dr Ian Oppermann is the NSW Government’s chief data scientist and CEO of the NSW Data Analytics Centre. Ian is a thought leader in the digital economy and is a regular speaker on big data and the impact of technology on society.
Data governance, or data sharing policy, is fundamental to smart city strategies. I had the opportunity to interview Dr Oppermann on this and related topics.
The conversation showed how data policy is key to issues such as addressing domestic and family violence. We think it’s a must-read for those who want our cities to be smarter. The interview will form the next three blogs – here is the first.
Steve Mackay (SM):The Data Analytics Centre describes its role as being to build analytics capabilities in order to improve health, safety, social, economic and environmental outcomes for citizens. Why isn’t that a relatively simple job of grabbing some of the already very advanced analytics tool-kits out there and finding out which one is most fit for purpose?
Ian Oppermann (IO):A big part of what we do is to try to get government agencies to think differently about the challenges they are facing. All agencies use data in one form or other, many have analytics capabilities. What is new is the ability to access far more data sets than ever before, and with this comes the ability to ask much bigger, more powerful questions. We also help agencies reframe their analytics questions to line up with the outcomes they are seeking to deliver.
With data from across government, you can ask questions normally beyond reach. Problems such as domestic and family violence are notoriously difficult to address without a joined-up view of what is happening. Knowing more of that journey or child / family / household as seen through data and being able to identify risks as seen from education, health or police means services can be redesigned and investment priorities reconsidered.
The analytical techniques we use are modern machine-learning techniques and this is new for some agencies. As the techniques gain broader acceptance, they will be much more widely taken up directly by agencies.
SM: We hear the term ‘data governance’ (or open or shared data) a lot – what exactly is it and what role does it play in building an analytics capability?
IO: Data governance is about understanding and managing the risk associated with the data we are using. Knowing how ‘personal’ data is, and how sensitive it is, forms the basis for the frameworks we use from initial collection to analysis, outputs generation and long-term storage of data. In some cases data has very low risk and can be made open. In some cases it is very personal (even when de-identified) or very sensitive, so great care is taken at each steps of the analysis process.
SM: If NSW develops a governance model which satisfies stakeholders here, aren’t other states and territories going to adopt a ‘not invented here’ stance and develop their own model?
IO: The challenges associated with data sharing are largely the same in all states and territories in Australia. The challenges also have international significance, as we are seeing through the work with Standards Australia. Privacy preserving data sharing has been a particular focus of the work we have been leading and it has been supported by representatives of every part of Australia. If we jointly develop a good solution, there is no reason to think other parts of Australia – and other countries – won’t use it.
SM: The challenge of data governance seems to have so many dimensions. Personal privacy versus open data-sharing; the ‘five safes’; rights of the people versus. that of the state; of those with less power versus those with more; of automatically generated ‘M2M’ data versus individually-entered social media data, and so on. Do you ever despair of finding a single comprehensive governance model? Do you hope to have a single model used by all free economies? Is this realistic?
IO: Despair? Only occasionally….
The challenges of data sharing are many. They are technical, cultural and societal. There is no one answer for every type of data and every use case.
We started off trying to tackle the challenge in 2016 when we asked: “why don’t people share data?” We ran a half day workshop and I thought, by the end of the workshop, we would have it all sorted. Almost three years later we are getting closer but still have a long way to go. We have now had 17 workshops, delivered two technical whitepapers and are about to release a report which describes the outcomes of some innovative experimentation. We have also kicked off international standards work with a host of other countries who have similar challenges.
We are getting a lot closer to understanding which parts of data sharing are subject to technology, which are subject to commonly accepted processes and which will remain in the domain of judgement. The parts supported by technology can be standardised, the parts supported by process can also be supported by standards, but the implementation or ‘settings’ for these processes will likely differ depending on what is societally acceptable.
Author/Interviewer: Steve Mackay
For a video of Ian Oppermann’s views: https://vimeo.com/193024310
