The trip to this workshop is part of a mentoring program in which Vicky is participating
Among the work carried out by the PhD students of the EC3 group is the initiative to participate with other colleagues in various activities, such as the one our colleague Vicky undertook by collaborating with two other PhD students on an interdisciplinary project that originated within RESPECT back in 2023. This is the second time Vicky has ventured into the world of conferences, leaving us with this brief chronicle of her experience:
Last week I attended the Workshop on The Use of Generative AI for Science of Science and Higher Education Studies (AISci), a scientific event organized by members of the Institute of Communication and Public Policy from the Università della Svizzera italiana, and held in beautiful Lugano, Switzerland. Together with PhD students Carolina Coimbra Vieira (Max Planck Institute for Demographic Research) and Elena Chechik (Europa-Universität Flensburg), we presented our ongoing work titled «Exploring ChatGPT as a tool for automated classification of US authors’ positivity perspectives on Russian topics». The Workshop was a really valuable opportunity for us, because learning about several different developments involving LLMs and discussing with experts in the field allowed us to clear doubts and overall improve our project.
" the Workshop gave us the possibility to connect with researchers and students from many different countries and institutions, and learn more about some very interesting ongoing projects that address Science of Science issues from the applied perspective of LLMs"
We started working on this collaborative project with Carol and Elena in the context of the RESPECT Fellowship, a mentoring program led by Dr. Cassidy Sugimoto that, since March 2023, brings together a cohort of interdisciplinary PhD students from all over the world. Our main line of research is related to the influence of bilateral relations on science, where we intend to study the case of US research on Russian topics partly through the implementation of ChatGPT’s annotation capabilities. AISci turned out to be the perfect event to present for the first time our methodological proposal, which combines human tasks with their large-scale automation through AI, receive feedback and, ultimately, plan the next steps of our development. At the same time, the Workshop gave us the possibility to connect with researchers and students from many different countries and institutions, and learn more about some very interesting ongoing projects that address Science of Science issues from the applied perspective of LLMs.
All in all, we left Lugano with a much clearer view of Generative AI’s capabilities as well as limitations, its great potential for the Social Sciences when used correctly, and the precautions we must take to avoid forcing its application to the development of tasks for which LLMs were not designed. I’m personally very grateful for having had the opportunity of attending AISci thanks to the organizers’ funding, acquiring a deeper understanding of these fast-paced technologies that are increasingly impacting all fields of knowledge, and visiting such a gorgeous little city by the lake’s bay where you feel immerse in Italian culture despite being in Switzerland.