Dr Walid Magdy, Lecturer in the School of Informatics, joins a panel of experts for Test Tube to YouTube – an examination of the behind-the-scenes factors that influence our social media feeds.
What can the audience at your event expect to see/hear/experience/learn?
People coming along can expect to learn something about the meaning of data science and its application in understanding and predicting human behaviour.
What do you hope that the audience will get out hearing about your research?
That social data from websites – such as YouTube, Twitter, and Facebook – can be analysed to reveal details of people’s interests and behaviours.
Why do you think your research is important and/or necessary?
Our research has many applications, for example in politics – such as understanding voter intent in elections; in the financial world, to enable targeted advertising, and in understanding social behaviour.
What might surprise people about the technical side of social media?
It is always surprising when we gain insights into how people behave and understand what motivates them.
What are the most technically challenging aspects of your work in predicting behaviour/ analysing content of social media/detecting political bias?
The coverage of data is the main challenge in my research. Social media is not representative of entire communities on the ground. Working with this problem in mind makes it challenging to get conclusive results.
You’re not a political scientist, but political or social inferences might be drawn from your results. What are your thoughts on this?
Data science can analyse large amounts of data, which until now social or political scientists less familiar with processing . By working together, we can perform amazing analyses which were not feasible before.
Is it possible for social media users to increase the chance of a post going viral and, if so, how?
There has been some research on this. However, there is no definite answer on what can go viral - it is usually unexpected. For example, why did the song Gangam Style go viral – was it really so good? Why did the game of water bottle flipping become an internet sensation and then fade in popularity? Why did one particular photo of a dead Syrian child gain more attention than hundreds of others? These are all questions for which much research is needed to answer.
Why is it important to you to take part in the Edinburgh International Science Festival?
Social data scientists work all the time with users’ data to understand human behaviour and community trends. Participating in the Festival can be like a payback to the community, to teach how their online social interactions are used by scientists to better understand the world and create useful applications.