Department team wins UN PET Lab Hackathon

This week a Department of Computer Science team took part in the United Nations Privacy Hackathon and placed 1^st out of 196 teams (325 participants) worldwide! Our team consisted of Manuj Mishra, Vishal Ramesh, and Konrad Kollnig. Manuj and Vishal are each pursuing the MSc in Advanced Computer Science while Konrad is a 4^th year CS DPhil student with a focus on privacy.

The competition included university teams from Harvard, Zurich, and Toronto as well as teams from industry, start-ups, and national agencies including the UK's own Office for National Statistics, who came 3^rd. The competition ran over 4 days from the 8^th to the 11^th November and had physical locations in Yogyakarta, Boston, and Zurich. The Oxford team took part remotely.

The challenge involved extracting valuable insights relating to refugees in Kenya during the COVID-19 crisis – such as the probability of them receiving government aid or the probability of them wanting to return to their home country. To do this the teamhad to make use of sensitive information collected in surveys by the UNHCR (United Nations High Commissioner for Refugees).

The competition was similar to other data science challenges with the exception that participants had to make use of state-of-the-art PETs (privacy-enhancing technologies) like differential privacy, secure enclaves and synthetic data to maintain the privacy of individuals.

As part of their prize, the team will be presenting their work at the 2023 UN Big Data Conference and will be writing a blog post with the organizers. Team members also won an iPad each. Sponsors of the competition included UN PET Lab, United Nations, UNHCR, Swiss Federal Statistical Office, IBM Research, OpenDP, OpenMined, Microsoft, and Amazon Web Services (AWS).

Shown in the team photo below are (left to right), Manuj, Vishal and Konrad