Since mid-June members of the Cloudy team have been at the Alan Turing Institute, a new a new national centre for data science . We're based down at the British Library in London and are hard at work trying to analyse all the data you've given us!
Here at the Alan Turing Institute, or ATI as we know it, they are running eight-week summer internships for PhD students with skills in something called data sciences. What this means is that these students are really good at finding ways to investigate big datasets - and Cloudy is a BIG data set! So far we've received over 2.2 million symptom reports from 8681 participants since January and we are grateful to you for continuing to report your symptoms as often as possible.
We're lucky to have three very accomplished students, Bertie, Konstantinos and David, who have agreed to use their skills to help us analyse the Cloudy data and who are already producing very exciting results. We are even beginning to write our first paper all about how you are using the app.
David, who is currently a PhD student at the University of Warwick, has been using different computer programs to try and identify patterns in data completion to understand what symptoms you are reporting and when. Once he has finished identifying these patterns, we will try to understand when and why people stop reporting their symptoms. This is important because it helps us to understand who takes part in our study and which patient groups we might be missing.
This is where Bertie (who you can see below pointing out some of his most interesting results!) comes in. Bertie is a PhD student from the University of Oxford and he has skills in text-analysis. He is using these skills to analyse the information you've sent us when you've given us feedback on the app. You might remember we recently sent you a push notification, for more of your thoughts on using the app? This was so that Bertie has more information to analyse. From Bertie's work we will try to identify reasons why some of you may not be managing to record data every day and why some people may no longer be taking part in the study – we will also let uMotif know some of your feedback to help them make the app better in the future.
Interns Bertie (far left) and Konstantinos regard their latest dataset discovery
Alongside the work of David and Bertie, Konstantinos (who you can see below with Bertie) has been able to use lots of exciting techniques to start examining the link between weather and pain. As you might imagine, we have an awful lot of data to manage and that means that a lot of the techniques we commonly use in research aren't suitable. We're lucky then that Konstantinos, who is a PhD student from the University of Edinburgh, specialises in 'machine learning', which basically means that he uses a variety of computer techniques to find patterns in big data sets, like Cloudy.
The results Konstantinos has produced help us to understand which items of weather data, such as temperature or pressure, we should pay attention to. He's also trying to work out what the time window is for the weather to affect your symptoms. It's really important to us that this work is done, so that we know how best to use your data in the future.
We have just 2 and a half weeks left of the internship, but over here at Cloudy HQ we are thrilled with the progress that our interns have already made. We've been learning a lot of very exciting things from David, Bertie and Konstantinos and very much hope to be able to share their results with you soon!