By Justin Ho
On 28th July 2018, I boarded the plane to Los Angeles, the place where I spent 11 days enjoying my year worth of sunshine and doing data science! I was selected to participate in the 2018 Summer School Series on Methods for Computational Social Science co-organized by University of Southern California‘s Information Sciences Institute and GESIS – Leibniz Institute for the Social Sciences. The summer focused on methods for analyzing and modeling textual data, featuring outstanding speakers studying different interesting research questions using computational methods and also students from all around the world. In teams, we also worked on small projects using the newly learned methods and under the supervision of the invited speakers.
During the opening session, I learnt that there were more than 280 applications and only around 30 were accepted. I was thrilled to be one of them, but my impostor syndrome immediately kicked in. For the small project, I was allocated into the team to work on the Twitter corpus collected during the Grenfell Tower Fire, a tragic event that happened in June 2017, with my teammates (Apoorva, Olga, Steffie, and Yimei) under the supervision of Dr. Miriam Fernandez. I am familiar with social media, I am a PhD candidate doing social media analysis afterall, but it was first time that I worked with such a gigantic corpus. There were 12 millions tweets and retweets and when working with such a vest volume of data, one simple operations could take hours and one wrong click would kill the programme. To me, it was as much a challenge for programming efficiency as a trial of modelling techniques. After a week of struggling, arguing, programming, and writing, we finally finished the project.
Looking back, it has been a fun, exciting, intense, but fruitful week. I learnt new knowledge, met interesting people, won an award in the mecca of data science, California, and most importantly, I had a lot of fun and gained a lot of weight!