Ajay Patel

Victim Assistance Using Large Language Models

Under guided research from Professor Cantay Caliskan and Ezgi Siir Kibris at the University of Rochester, we investigated the usability of large language models to conduct sentiment analysis as an alternative to traditional survey-based methods in disaster-affected regions. This included prompting various LLMs with possible post-disaster events and analyzing changes in sentiment. Our full research was published at the Social Science Research Network here.