In this blog interview, we ask Merve Alanyali to share her experience of hacking happiness, and to tell us more about her data science research at the Turing. 

Merve is a Turing enrichment year student from University of Warwick and is now organising a new Data Dive event to be held at the Turing with Turing Faculty Fellows Tobias Preis and Suzy Moat.

The hackathon topic was happiness. A topic often discussed within the humanities rather than a subject of scientific investigation. How did you feel about approaching it through data science?

Doctoral students Merve Alanyali and Chanuki Seresinhe at the Hacking Happiness Hackathon organised by the Digital Catapult and the Royal Society. Franck Fourniol‏ @FrancknScience

What I am dealing with in my research is not that different. I am not dealing with happiness but I am trying to quantify human behaviour and happiness is an emotion that can be reflected in human behaviour.

To think about how we would approach this subject from a data science perspective we should probably start from something more general: defining what data is. It may seem a cheesy sentence but data is life. You do everything based on data. Even when you describe what you are going to eat for dinner tonight, you think about what you have at home. For example, the ingredients, which is data; you may think what you had last night, that is again data: your historical data. We are not aware of it; but most of our reasoning is automatically based on data. To think about data from a general perspective, you don’t need to be a scientist.

Data is raw form of information. You can process and extract some information from data.  In data science, we use the scientific methodology to extract that meaningful information, to interpret signals from the data set. That’s my very general description of data science.

In what sense is hacking happiness not that different from what you are doing?

The field I am working on is called computational social science or social data science. We analyse large quantities of data generated by the everyday interaction with technology based devices, like phones, computers, etc.

With this, we can analyse social processes as they unfold or we can improve predictions on human behaviour by identifying behavioural patterns.

Personally, I am quantifying human behaviour using online images. This could be data attached to online images, such as geotags, timestamps or text data for example pictures tags, title or description.

My PhD is made by different subtopics and I am working on various case studies. As an example, one case study is detecting protests by analysing online images of worldwide protests. If you want to hear more about my work you can watch my Turing short talk video on YouTube.

Another example of computational social science could be Chanuki Seresinhe’s work: her research focuses on how scenic environments affect people’s wellbeing.

In this context, the idea of measuring happiness or hacking happiness wasn’t something novel to me.

Merve's blog

In the instance of the Hackathon, which idea did your group develop to measure happiness? 

We came up with the idea of a web application, called ‘Boroughmeter’, that would tell you in which part of London you would be happier if you had to move to a new area. For example, in searching for a place to live, you can select among a set of criteria and given your selections the application would suggest the best area for you to live.

In order to build such an app, we used data from a mobile app called Mappiness. This app aims to quantify your level of happiness. It buzzes a couple of time a day and asks how happy you are and grades your level of happiness on a scale. You may ask whether this is how you can measure happiness.

My idea of happiness could be very different from your idea of happiness. But one way of measuring it could involve surveying people and asking them how happy they are or how happy they think they are. We could gain a ground measure, just by directly asking these questions and asking people to rate their feelings from 1 to 5.

Although we built a working demo during the hackathon, there is still a lot of work to do and room for improvement but this is the nature of the competition. You come up with an idea and make it work in a very short amount of time while working with a new and diverse group of people.

Would you recommend participating in a hackathon to your colleagues?

Yes definitely. It is a great experience working with a people from different fields towards achieving a common goal: coming up with an idea and making it work in a limited amount of time.

I am organising a new hackathon with Turing Faculty Fellows Tobias Preis and Suzy Moat, directors of the Data Science Lab at Warwick Business School, University of Warwick. We are also very excited to be collaborating with the new Office for National Statistics Data Science Campus for this event.  Our data dive will be held here at the Turing Institute, with a focus on urban analytics. Stay tuned for more information!

Which are the most valuable aspects of your experience here at the Turing?

I really like the mixture of universities at the Turing; meeting new people from different universities working towards a similar goal. It feels like we have joined a new community on top of the single universities we belong to, which I think is quite valuable. I also really enjoy the events that Turing is organising and the location advantages; being in a very central position in London allows me to attend other interesting events just after work.