Data Bites 58: data visualisation, automation and APIs

Data Bites returned from its summer break on 18 September with three brilliant speakers on data visualisation and automation, a government-developed tool to match companies across different datasets, and APIs.

Our fourth speaker – Andrea Barry, on Youth Futures Foundation’s new data dashboard – was sadly unable to join us, but we hope to hear about the project at Data Bites soon.

If you’ve not been (or seen) before: Data Bites is a monthly event showcasing how data is used in and around the public sector. Four speakers get eight minutes to present – there are eight bits in a byte, hence eight minutes in a Data Bite – followed by eight minutes of audience questions. September’s event was the ninth Data Bites hosted by Public Digital, after 49 events at the Institute for Government.

You can now sign up in person or online for the next Data Bites, an AI special on 23 October.

Ahmad Barclay, Principal Data Visualisation Specialist at the Office for National Statistics, talked us through some of the ONS’s work on dataviz, particularly around the 2021 Census.

Ahmad started by looking at the different types of data users and what they need; building up layered narratives through maps and charts; how a statistical technique called Moran’s I helped; the importance of starting with a local narrative; and how to automate some of the processes. 

Leonardo Mazzone, Senior Data Scientist at the Department for Business and Trade, presented Matchbox – a tool developed by DBT to help match the same companies across different datasets. You can find out more in the team’s repository on GitHub, by taking a look at Leo’s slides, or checking out Leo’s answers to our questions, below.

Can you summarise your project in a few sentences?

Matchbox is an open-source tool for matching records across datasets. It helps teams deduplicate and connect information from different sources in a secure, collaborative, measurable and iterative way, enabling better analysis and improved digital services.

What problem are you trying to solve?

Government departments often hold fragmented data about the same entities (like companies). This fragmentation makes it hard to deliver seamless services, effective government interventions, and data-driven policies.

What difference will it make to citizens?

By simplifying the linking of data, Matchbox can help government services feel more joined-up and responsive, reducing the need for citizens and businesses to re-enter information and enabling more tailored, while having their needs met more efficiently. Citizens will also benefit from better policies and government interventions, supported by higher quality and richer data.

How will you know when it’s succeeded?

In the first instance, when it becomes standard for analysts and developers at the Department for Business and Trade to use Matchbox data to implement a new feature, service or dashboard, and they can do it in a day or less. Another organisation adopting Matchbox and contributing back would also be a significant success.

What are the main hurdles?

To reap the full benefits, there needs to be engagement and buy-in from multiple organisations. Technically, issues such as diverging data models or versions of the truth between different services.

If you could change one thing about data in government, what would it be?

Stronger incentives for the re-use of tools, methodologies and data, to discourage different organisations solving the same problems in isolation.

James Freeland, Senior Data and Security Architect at the Government Digital Service, presented about GDS’s API programme. You can read more about it in a blogpost published by the team, in James’s slides, or in his answers to our questions after the video. 

Can you summarise your project in a few sentences?

The project aims to improve API Maturity across dept and public sector organisation; foster improved developer experience, enablement and collaboration; make it easier for users to access a central catalogue in order to find, discover, and search API; standardise API practices and reusability of API artefacts across gov in order to improve interoperability and data sharing; and provision of guidance on technical standards and support to users across gov. We explored these problems through a pre-discovery; structured user research led discovery; and tested findings and recommendations in Alpha; and eventually decided a centralised API Hub for users across gov will be required to help improve API maturity, standards and accelerate API uptake across gov

What problem are you trying to solve?

• Fragmented API catalogues causing duplication issues

• API documentation is of poor quality

• Lack of a standard approach to implementing APIs

• Uncertainty on how to make APIs secure

• Value of APIs not fully understood by all users (business & technical)

What difference will it make to citizens?

• Better public services on data use and policy making which will have an impact on citizens lives
• Improve citizen data access 
• Sharing of data will make it easier and faster for local government, charities, researchers, and citizens to access data that they need
• Access to apps make it easier for citizen to access data in a secure way

How will you know when it’s succeeded?

• Improved API Maturity across gov
• Increase in uptake in API products of gov departments and public sector org exposing their services to API
• Metrics to show business value of API

What are the main hurdles?

• Legacy systems sometimes can complex and not easy to access
• Reluctance by data owners/organisation to share/or expose services to API or depart from manual approaches they are used to
• Business users not knowing the true value of using APIs and benefits it offers in long term savings in costs and time
• Cultural shift and mindset change for new users.

If you could change one thing about data in government, what would it be?

Make data sharing easier by providing the right digital tools to this in safe and secure manner.