Digital

Data Science Accelerator launches for 2022

October 7, 2022 by Stewart Hamilton No Comments | Category Data, Digital Scotland

Blog by Joyce Dalgleish, Business Change Manager, Scottish Public Sector Analytical Collaborative.

After months of planning, we were delighted to finally be able to bring the Data Science Accelerator cohort of participants and mentors for 2022 together with The Data Lab at our recent launch event in Edinburgh. This year marks the fifth edition of this innovative capability building programme and we are delighted to welcome 16 applicants and volunteer mentors onto the accelerator.

Chief Statistician, Ally McAlpine, opened proceedings by sharing his own experience of the first accelerator in 2018 and how it had been instrumental in helping to modernise Scottish Government’s agriculture statistics. The accelerator had provided the support for his team to transform a labour intensive data collection process using telephone and questionnaires, into one that used satellite data and a machine learning algorithm that can recognise the crops growing in 400,000 fields across Scotland from radar images.

Ally thanked “the applicants for being bold, for putting your hand up and saying I see what needs to improve, the mentors who are willing to share their hard won experience to nurture that enthusiasm, and to the organisations who know that to change we need to create an environment where ideas can flourish”.

We then heard from each of our applicants about their projects, which cover everything from improving road services to accurately analysing the sentiments behind social media posts about Scotland.

First to speak was James Bradbury of Registers of Scotland who is developing a tool to create text files from scanned legal documents that will automatically extract specific information and present this in a more human readable way. The short-term goal is to build a faster manual review system with a longer-term goal to totally automate the extraction. Shortening the time this process takes will benefit stakeholders as well as reduce the administrative burden of manual checking. James will be supported by his mentor, Tom Wilson, from Scottish Government.

Joining us remotely were Bruce McMaster and Ryan Glass from Dumfries and Galloway Council who have taken as their challenge the need to make better use of existing local authority data to improve the planning and delivery of the roads service. Bruce and Ryan will be working on this with their mentor, German Priks of NHS National Services Scotland.

The project submitted by Maria Ten Sierra from NHS Lothian will explore the use of data collected in free text fields in hospital records. The aim is to find a way to make this information more visible to consultants and GPs to assist with clinical assessments for referring and triaging. Sam Hunter from National Records of Scotland will be mentoring this project.

This year brings 3 projects from Public Health Scotland.

The first of these is from Bob Taylor who is developing an interactive agent-based approach that will allow modelling of different disease spread scenarios through an interactive Jupyter notebook. Bob is aiming to allow users to explore the effect of modifying infection parameters and environments on disease spread. He will be working on this with his mentor, Simon Rogers from NHS National Services Scotland.

Eleftheria Vasileiou is seeking to automate the production of Blood Borne Viruses and Sexually Transmitted Infections surveillance reports as well as provide authorised users in NHS boards with a management information dashboard. These users will then be able to interact with their data locally. Nikola Fanton of Scottish Government will be mentoring Eleftheria on her project.

Rob Gibbs and Steph Taylor are building a proof of concept dashboard that contains information on medicines prescribed by GPs and pharmacists for mental health conditions across NHS Scotland. Their aim is to make this data more accessible so that the users can create their own visuals for their organisation or area of interest and they will be supported by their mentor, also from Public Health Scotland, Rita Nogueira.

The Visit Scotland team, Carly Walker and Tutku Kose, will explore how to more accurately classify sentiments expressed in social media posts. They want to create a bespoke sentiment analysis for Visit Scotland that will pick up on slang and colloquialisms as well as Scots and Doric. Carly and Tutku will be mentored by Vyron Christodoulou from The Data Lab.

As well this fantastic list of projects, we’ve also had a bumper crop of mentors and it’s great to see that several of our volunteers have previously been participants in the accelerator. It’s been our ambition to create a learning cycle where participants return to the programme to mentor others and of course continue to develop their skills the process.

Our lead mentor for the accelerator is Joanna McKenzie from The Data Lab. Joanna helpfully reminded us that the accelerator is really only a very short time – it is just 12 days, so it’s important to focus on delivering a minimum viable product not a final, polished article. These will all be iterative, agile developments and as Ally McAlpine’s experience showed – these projects do take time to realise their full potential.

Our closing speaker was Alan Howie, Chief Data Officer for Registers of Scotland. Alan spoke about the managers experience and how beneficial it is for an organisation to have staff go through the programme. He described the accelerator as a way of “revealing the talent and potential within your staff”. And Alan had a fantastic example of that from his own organisation. Marguerite Le Riche, a former accelerator participant who is now their lead Data Scientist.

With the accelerator underway, we want to take this opportunity to thank all the participants, mentors, their managers and their organisations. Thank you to Scotland’s publicly funded data innovation centre, The Data Lab and to the Scottish Public Sector Analytical Collaborative (Scottish Government, Public Health Scotland, Registers of Scotland and National Records of Scotland) for their support. We are all looking forward to seeing what exciting outcomes our projects will deliver.

If you’d like to find out more about the accelerator, please get in touch with us at: DataScienceAccelerator@gov.scot.

Tags: Data Science Accelerator, Official statistics, Public Sector, scotland, Scottish Public Sector Analytical Collaborative (SPACe)

Comments