The NatStats 2013 Conference Communique.pdf is available in pdf.
The Australian Bureau of Statistics (ABS) hosted its third NatStats conference at the Brisbane Convention & Exhibition Centre in Queensland on 12-14 March 2013. Over 350 delegates attended the two days discussing how the game of statistics is changing, and the opportunities presented by big data to drive productivity and build a strong and vibrant nation.
The theme of the conference was “A better informed Australia: The role of statistics in building the nation”. Discussions focussed on how big data (defined as administration, transactional, Web 2.0, sensory network, international and emerging datasets) can be used to inform decision making in the social well-being, economic and environment domains, as well as looking at the challenges of extracting knowledge from big data.
Defining big data
The conference recognised that big data means different things to different people, and throughout the sessions several interpretations were provided. That said, there was agreement that big data can be characterised by the ‘5 Vs’: high volume, high velocity, and high variety information assets that represent high value and veracity for users of statistics. These can include, but are not limited to, accessing administrative datasets and by-product data not originally intended for statistical use, to harnessing the informative content from social media and sensory networks.
Ros Harvey, Director of Tasmania’s SenseT Project said during her presentation, “We are awash with data. The challenge is how we distil its value.”
Discussions highlighted the value of big data in providing opportunities and benefits for a range of sectors including government, community and the private sector.
How the data game is changing
The presentations on the first day highlighted that big data can provide opportunities to enhance our picture of the everyday experiences and activities of Australians, individually and as a community. It also highlighted the opportunity for businesses to use this information to adapt and innovate in the 21st century.
However, to extract the maximum benefits of big data and seize on these opportunities, we need to address some challenges in order to convert big data into knowledge for decision making. The skills set needed by those involved in the collection of the data, the data scientists engaged in extracting knowledge from the data and by users in making best use of that knowledge will need to continue to evolve from current levels to meet the demands of tomorrow.
As noted by Alan Smith from the UK Office of National Statistics on day two of the conference, “The emergence of big data in everyday society is not in itself the current revolution, but rather the revolution is in the recognition of the potential value of big data assets”.
Throughout the conference, there was a strong recognition that the ‘big data deluge’ is changing the information game and that we need a focussed agenda in order to harness its potential across private and public sectors, nationally and globally.
Where are the big data opportunities?
During the afternoon of the first day, delegates discussed the big data opportunities in more detail, focusing on how they can improve policy development, service delivery and productivity. The areas of Education, Health and Relative Disadvantage were identified as key opportunity areas where big data including, longitudinal data, has the potential to help answer complex social wellbeing questions. In the area of productivity, measurements of human capital were voted as most likely to benefit from exploiting the potential of big data. Meanwhile, establishing stronger links between bio-physical, economic and social data sets were seen as the biggest area of potential for utilising big data in supporting environmental policy development.
The opening sessions of the second day provided the opportunity for delegates to consider some of the challenges in converting big data into knowledge. The rapid evolution in data visualisation techniques and the growth in demand for them as a contemporary tool to distil the knowledge from data were highlighted as a key enabler that the statistical community will need to continue to invest in to extract statistical stories and understanding from big data.
The importance of allowing the data to be analysed and to work with key partners was discussed. One example discussed was the role of the media as ‘retailers’ of the knowledge inherent from statistics as a critical link in ensuring that the value of statistics is fully realised. This reminded us about the importance of ensuring that audiences can relate and resonate with the information being conveyed beyond the facts of any numbers being presented.
What are the big data challenges?
Delegates then took the opportunity to consider some of the challenges associated with big data in more detail and consider priorities for future focus.
Maintaining confidentiality and privacy will always be a primary criterion for ensuring ongoing access to these valuable assets. In considering confidentiality and privacy issues, community trust and restrictions imposed by legislation were identified as the most significant challenges.
Currently other key barriers to extracting maximum value from existing big data assets include the costs or lack of funding associated with data collection and both people and technology capability shortages. In considering knowledge and capability issues, the convergence of analytical, statistical and technical skill sets into ‘data scientist’ capabilities was raised as a key future need. Enhancing information management capability and the statistical literacy of end users of statistics were also identified as key priority areas.
How can big data enhance Australia’s statistical system?
The expectations and complexity of users’ needs are increasing, and traditional statistical systems often fail to deliver the answers needed, or deliver them in sufficient detail or timeliness, the outcome being increased uncertainty in decision making. The conference highlighted the wide range of opportunities and challenges that big data provides for data custodians, producers and users of statistics.
There was strong support for the need for a combination of approaches in addressing confidentiality and privacy issues, including: more effective community engagement on the benefits of data integration, investment in standardised tools to support confidentialisation of data; and adoption of agreed data integration principles across sectors.
There was also recognition that big data provides an opportunity to enhance the broader Australian statistical system. Priority areas highlighted by delegates included prioritising investment in the nation’s most essential statistical assets, enhanced governance arrangements for the Australian statistical system, improving access to big data, and further embedding mathematics and statistical science into Australia’s education system.
The Australian Statistician closed the conference by stating that realising the potential of big data to shape our future will take time and we will need to collaborate and work together as a statistical community to succeed. It will also require investment from governments over time if the potential of 'big data' is to be realised for societal benefit.