Drilling down for data value

Data might be the new oil, but government agencies need to understand how to utilise the tools available to extract the value from their data assets

Tags: Big dataData analyticsForrester Middle EastMicrosoft CorporationOracle CorporationSAPTeradata (www.teradata.com)
  • E-Mail
Drilling down for data value Most governments have established data projects, but the tools available are evolving.
By  Mark Sutton Published  April 30, 2018

Data has become the buzzword for many government projects, indeed across the whole of the ICT sector, CIOs are being joined by data scientists and chief data officers as organisations try to harness this new wave of benefits that are promised through better understanding, analysis and measurement of data.

 Leaders such as Dubai are well on the way to the data-driven future, but many public sector entities are still faced with unresponsive paper-based records, unconnected silos of data housed in out-dated repositories, or just simple spreadsheets as tools try to mine the value from their data. As digitisation progresses however, it is clear that all government organisations need to adapt to data as the fuel for creating value, developing insights and designing new services.

The value of analytics and data to government is already well-established, there is even a growing movement around ‘Data for Good’, where developers, data scientists and other experts across public and private sector and academia are coming together to use analytics and data to address social challenges.

In a report from Forrester concerning Data for Good, analysts Jennifer Belissent PhD, and Elizabeth Cullen point out that: “Government agencies, non-profits, and intergovernmental organisations are not blind to the emerging trends in data science and prescriptive analytics. Policymakers, advocates, CIOs, and technology organisation teams appreciate that advanced statistical modelling can uncover insights that enhance citizen engagement, streamline bureaucratic processes, and address pressing social issues like poverty, environmental degradation, homelessness, hunger, and discrimination.”

The Forrester report highlights how data analytics, especially cross-agency or using a mixture of government-held data and publically available data, is producing some unexpected benefits and services. Many of these projects are already well underway. For example, the city of Chicago has a program for predicting restaurant hygiene. The city’s data science team developed a predictive model that prioritizes restaurants for inspection by analysing variables from online data sources including social media and weather and internal records, such as past inspections, business licensing and so on. In the trial phase conducted in 2014, inspectors were able to uncover critical violations of restaurant codes nearly a week earlier than previously capable.

Other projects include the state of Indiana developing a better understanding of infant mortality by using SAP Hana to analyse 17 data sets from five agencies and four public sources; or the US Department of Veteran’s Affairs developing a model to predict suicide among veterans through analysing 61 variables in patient health records.

In the UK, Oracle ran a prototype project with the National Health Service to analyse patient health records and drug prescription data using machine learning to identify £100 million in savings in the first three months of operations alone.

Data analytics cut across many different areas of government engagement as well. In the Middle East, Kuwait’s Directorate General of Civil Aviation (DGCA) utilised a solution based on the Microsoft Azure cloud and Power BI analytics to analyse every facet of airport operations, from air traffic control to facilities maintenance, to deliver dashboards that can be used by many job functions to create new efficiencies. In the UAE, Expo 2020 Dubai has partnered with SAP for a wide-ranging project to deliver real-time insights for the expected 25 million visitors that the Expo will bring to the city.

The Forrester report highlights some of the steps which entities need to take to realise their data ambitions, such as the need to foster and retain skills, build their internal capabilities and scale them, and the need for both leadership on data from the top and for data projects to deliver insights that can be acted upon in a timely manner.

A major hurdle for public sector projects however, is lack of access to the right tools. Based on the results of the Forrester Data Global Business Technographics Data and Analytics Survey, 2017, the report authors note: “Most government organisations lack a strong data architecture to support analytics. As a result, it’s not surprising that barely half of government data and analytics decision makers reported being able to access the necessary data to generate insights in a timely manner, compared with almost three-quarters of their peers in other industries. Innovative analytics initiatives often require data from external sources, yet only 35% of government decision makers report expanding their ability to use data from third-party sources, compared with 59% of their nongovernment peers.”

Government requirements for data tools have moved beyond just data repositories, however. Jamil Jeitani, managing director, Teradata Saudi Arabia, explained: “As government organisations become more sophisticated and responsive, it is critical to move beyond data collection and storage. The power of analytics makes it easy to harness an agency’s data, structured and unstructured, to reveal hidden patterns, trends, citizen preferences, previously unknown correlations and other useful actionable knowledge. Leveraging big data and analytics to uncover data insights are essential for helping the government organisation make critical decisions, faster.”

Savio Monteiro, Director, Software Solutions Group, Gulf Business Machines (GBM) said that government organisations are looking for solutions which will give them better insight into data and the capability to equip non-specialist staff with the ability to work with data.

“Most of our analytics projects lie around implementing solutions for clients that provide them with reporting and analysis, planning and budgeting, enterprise insight analysis, big data, enterprise data warehouse, and predictive analytics.

“Governments are looking for solutions that provide them with the tools to make better decisions and streamline their operations, both in terms of ease of service for the citizen and cost. They want better access to all of their data and to be able to examine it from every angle and gain valuable insights. They like the autonomy these solutions provide, as they no longer require a separate team to access or examine the data. With the power of analytics, governments can better understand and interpret the data they already have,” Monteiro said.

There are also different demands from different parts of the organisation. Necip Ozyucel, Cloud and Enterprise Business Solutions Lead, Microsoft Gulf, explained that the ‘business’ stakeholders are looking for adaptability, and the ability to respond quickly to new requirements and to deliver clear, actionable and meaningful insights from data. At the same time, the IT function will be concerned with governance and delivery, and how they can deliver a secure, scalable and consistent platform.

“The optimum solution, of course, will come from the solution provider who has identified the push-and-pull nature of these two viewpoints, and offers a platform that satisfies both governance and agility requirements,” Ozyucel noted.

He also said that government data projects will often require tools and solutions that can function across different domains: “Most private-sector companies operate in just one industry. A healthcare provider will require an analytics platform that deals with one broad type of data. An airline or automobile manufacturer will have very different requirements; but again, within a specific industry, the type of data will seldom vary wildly.

“Governments not only need to warehouse data across a daunting range of interests — administration, taxation, economic development, national defence, policing, education, healthcare and more — but different governments have different priorities. Even in the GCC, where all governments are pursuing long-term, national economic vision programmes, the data they prioritise may vary. The challenge is to present public entities with the most customisable tools possible, so they can build the analytics platform they need to meet their individual nation’s challenges,” he said.

Cloud platforms, such as Oracle’s Analytics Cloud (OAC), enable this speed and scale, and also help agencies to make the best use of all data, no matter where it is, she said. Provision of different options for analysis mean that different users can easily be assigned the right level of tools to enable effective analysis, so that more people can be engaged in analytics projects.

Another development in analytics is the introduction of machine learning and AI both to provide new capabilities and to support users of analytics solutions. Rizwan noted that AI-powered apps can perform higher-order tasks typically performed by knowledge workers. This can manifest in diverse areas even such as using data-insight driven smart bots to encourage individuals to quit smoking, through constant support delivered via smartphone, she added.

Machine learning can also support and empower the users of the analytics solutions. SAP has its Leonardo Machine Learning Foundation, which connects developers, partners, and customers via the cloud, to enable collaboration on applying machine learning to data, without the need for the organisation to develop the highest level of skills inhouse. Capabilities such as ‘out-of-the-box algorithms’ also allow organisations to get to grips with their data more easily.

Dr Hichem Maya, Head of Industries, SAP MENA commented: “Equally important as adopting data analytics and machine learning solutions is that government organisations need to train their staff on using data analytics, artificial intelligence, and machine learning.

“Not everyone has to be a data scientist. But with more possible insights than ever before, more staff do need to be familiar with how to analyse the data, and work with developers to launch mobile apps that can deliver these insights in a way that the general public can understand, and that could be monetized by partners.”

Machine learning can also help to discover patterns and relationships that are not expected, Microsoft’s Ozyucel added. Machine learning is often put to use in trying to find connections in ‘labelled data’, historic data where the outcomes are already known, he said, where the machine learning algorithm tries to find connections between pre-existing data and outcomes, in a process called ‘supervised learning’.

Governments are also interested in how machine learning can be applied as ‘unsupervised learning’ which works on data with no labels, presenting results in the form of patterns, interesting subsets of data, or analysing the whole structure of the data to find unexpected correlations.

“Machine learning offers exceptional value to analytics platforms. Its very nature means it can give answers to questions that even the most innovative decision makers would not think of asking, which is of great use to governments trying to leverage public resources more efficiently,” he said.

Add a Comment

Your display name This field is mandatory

Your e-mail address This field is mandatory (Your e-mail address won't be published)

Security code