Modern business intelligence software allows data and analytics to be shared to a larger audience – both internal and external to an organisation. It provides the benefit of a single point of truth of data, while enabling an analysis that is meaningful. But for organisational value, there needs to be trust in the system and processes that surround it. And with data often holding confidential or operational-sensitive information, it is important that stringent security surrounds it.
The Australian Institute of Health and Welfare (AIHW) produces high-quality reports, products and services providing insights into key health and welfare issues in Australia — used to inform policies and programs aimed at improving Australians’ health and wellbeing.
The data they use can include information provided by federal and state and territory governments, research organisations, as well as survey data containing individuals’ responses.
In providing their services, it is essential for the institute to adhere to strict privacy requirements. And this needs to be balanced with the need to ensure stakeholders, and the public, are better informed of health and welfare issues impacting them.
In developing a solution for improving data access and literacy securely, AIHW implemented an environment using Tableau Server.
Their process begins with creating data extracts that are have stripped any details that are considered likely to contain personal information and possibly breach privacy requirements.
Using Tableau Desktop, visualisations are developed and are published to an internal server in the development environment. After development, the visualisation is pushed into an environment for checking.
Any release of information must go through an approval process that includes checking that privacy and confidentiality of data is maintained. Once a visualisation is approved, it is moved into a pre-release area.
Approved visualisations are then transferred to the Tableau Server that exists on their external environment, creating an extra layer of security between data and output.
To support public delivery of visualisations, unlimited guest account privileges are available on the external Server environment, but they have also created a restricted embargo environment allowing visualisations to be publicly shared with access given to selected stakeholders and the media prior to their official launch. And this has no limitations on user numbers due to their core-based license.
The resulting data visualisation provide improved access to data and information on alcohol, tobacco and illicit drugs, health expenditure, obesity, palliative-care related hospitalisations and homelessness services – just to name a few.
The AIHW has recently reached a significant milestone with over 1,000 interactive visualisations made available to the public in under 18 months of implementing this environment. This provides an example of how the AIHW is supporting the Australian Government’s open data agenda and recognises the impact this can have in improving the health and welfare of Australians.
AIHW is the perfect example of how software and systems, combined with internal data governance models, can facilitate bridging the gap between security and public engagement.
Governance is required for tools as well as data
In thinking about governance and security, it is not just the tool – the visualisation – but also the data. Processes need to be established for linking and cleansing of data. Rules need to be established to ensure individuals cannot be identified. Metadata needs to provide information on the dataset, source and quality. And access needs to be provided to all those who need it, with the aim of having a single point of truth and avoiding data duplication.
For AIHW, cleansing exists in the process of creating data extracts before bringing them into the server environment. But the solutions also needs to facilitate data processing and management, to better power visualisations and business intelligence.
It is a challenge both public and private enterprise face in developing whole of agency data solutions.
For Cboe Global Markets, one of the world’s largest exchange holding companies, it is a challenge to build confidence in data sharing across departments and roles. The norm was multiple copies of the same dataset. Not only does this prevent a single point of truth, but the data quality can vary greatly depending on its age and what “cleaning” has occurred.
“One of the biggest concerns, even for me, with giving everyone access to data, is Shadow IT,” Lorena Vazquez, a Senior Software Engineer at Cboe Global Markets recently explained. “Everyone has their own copies of data, they use different terminology for the same dimension or measure, and it becomes an overall data governance nightmare.”
It is a common story. But moving away from processes that have long been cemented within an organisation is challenging. It is important that staff working with data see benefits – and through better data processing and management, they can get access to better data and more frequently.
Moving to a server based solution allowed Cboe to automated data cleansing and updates, improving the quality of data. Showing this was an important advocacy tool to move Cboe analysts away from their siloed approach to data.
Understanding the importance of governance in secure data access
While software and systems form a key part in the ability to deliver secure access to data, and make the process easier, a model of collaboration, review and approval needs to be a central component of the solution to ensure the right information reaches the right audience – as demonstrated by AIHW.
Buy in and participation is important to establish and maintain governance in these environments. Establishing a steering committee is an important start to this process. It brings together key players to establish a clear vision and framework.
This group can define rules, expectations and processes, including user roles, privileges and permissions, training and certification – and ways to measure success.
The support of IT departments are important in the successful implementation and configuration of software and systems. But as business intelligence enables data and analytics to be available widely within an organisation, a shift needs to occur in governance to bring policy makers and other areas of an organisation to the forefront, allowing them greater awareness and control on issues of access and security – and the ability to identify, report and resolve data issues.
For all organisations, secure access to data through business intelligence begins with software. But it requires the support of people and policies to succeed. As the AIHW show, capability and drive exists within Australia’s public service. And it is within the grasp of any Australian government agency with the mindset to improve the power of their data.