It gives us great pleasure to announce the winners of the 2020 OII MSc Thesis Prizes.  The Board of Examiners awarded one Thesis Prize for the MSc in Social Data Science and one Thesis Prize for the MSc in Social Science of the Internet, plus four ‘Highly Commended’ prizes.

Congratulations to Zo Ahmed, MSc in Social Data Science

Zohaib Ahmed

Thesis title: Tackling Racial Bias in Automated Online Hate Detection: Towards Fairer, Context-Aware Classification of Hateful Users on Twitter Using Geometric Deep Learning

Zo said, “I’m delighted to receive the OII thesis prize and very grateful to my supervisors, Renaud, Scott and Bertie, for inspiring and supporting me. The OII is a place which marries a range of different disciplines, from cutting-edge deep learning to critical discourse analysis. Being exposed to these multiple perspectives, and with the support of the brilliant people I met at the department, I was able to complete a research project that, though stubbornly difficult at several points, proved to be extremely rewarding.”

In congratulating Zo, Dr Scott Hale and Prof Renaud Lambiotte said, “We are delighted for this recognition of Zo’s work. His thesis is remarkable in its breadth and elegantly combines tools from machine learning and social networks in order to develop fairer hate detection algorithms. It was a pleasure supervising such an independent and skilled researcher.”

Dr Bertie Vidgen, Research Fellow at the Alan Turing Institute added, “Automatically detecting purveyors of hate is crucial for keeping online spaces safe and accessible, but it also presents a series of complex social and technical challenges. This work makes a real step forward in addressing them, presenting a detection system which is both higher performing and fairer than previous efforts.”

Congratulations to Claire Woodcock, MSc in Social Science of the Internet (part-time)

Claire Woodcock

Thesis title: Dr Bots: The Impact of Explanations on Layperson Trust in AI-Driven Symptom Checking App Diagnosis

Claire said, “I am deeply honoured to be awarded this prize. Studying alongside such a brilliant cohort, I am humbled my work was chosen. The OII provided an intellectually exhilarating experience in examining the societal impact of technology through a multidisciplinary lens. I would like to extend special thanks to my supervisors: Dr Brent Mittelstadt for his guidance on the philosophy and ethics of explanation, and Dr Grant Blank for his advice on quantitative methodologies and study design. Their practical and thought-provoking conversations opened my mind to combining approaches, enabling a holistic analysis. As a mature student juggling work, life and the impact of a pandemic, their unwavering encouragement certainly made this thesis happen. We’re entering a time where remote healthcare is necessitated and automation is seen as a tool to efficiency. I believe the success of this thesis underlines the importance of putting the human first in automated decision making.”

In congratulating Claire, her supervisors Dr Brent Mittelstadt and Dr Grant Blank said, “We are delighted that Claire has been awarded the thesis prize for her exceptional interdisciplinary study. New challenges posed by AI and automated technologies require careful interdisciplinary analysis to unpack and fix, especially in the context of healthcare and human-computer interaction. Claire achieves precisely that, combining an innovative experimental study design with insightful social and ethical analysis. Her dedication and creativity in the project took what is inherently a difficult phenomenon to study and made it accessible while generating key novel insights on how best to explain the behaviour of AI to people. We congratulate Claire for producing such a brilliant thesis and on her well-deserved prize.”

 

The Board of Examiners also chose to award four ‘Highly Commended’ prizes this year:

Highly Commended: Lisa Oswald, MSc in Social Data Science

Lisa Oswald

Thesis Title: Should we Talk to Climate Skeptics? Temporal and structural analysis of user attitudes and communication behaviours within a major climate skeptic forum on Reddit

Highly Commended: Marcel Schliebs, MSc in Social Data Science

Thesis Title: Understanding Social Distribution Networks of Chinese State-Backed Media

Highly Commended: Nate Zinda, MSc in Social Science of the Internet

Nate Zinda

Thesis Title: The Divorce of Physis and Techne: A Computational Content Analysis of Anti-Tech Extremist Organizations

Highly Commended: Asger Nim, MSc in Social Science of the Internet

Asger Nim

Thesis Title: Who sets the agenda? How journalists’ use of Twitter makes them vulnerable to computational propaganda