A Multimodal Analysis of Embedded Climate Change Media on Twitter

Project Information

ai, AI/ML
Project Status: Complete
Project Region: CAREERS
Submitted By: Gaurav Khanna
Project Email: sli@bryant.edu
Project Institution: Bryant University
Anchor Institution: CR-University of Rhode Island

Students: Jacob Dauphinais

Project Description

This project seeks to analyze video and image elements of Twitter/X posts in their association to climate change. The focus is on data pipeline to create textual descriptions using Large Language Models. More specifically the objectives are:

-Develop a methodology for video and image processing using open-source LLMs
-Test multiple open-source LLMs for effectiveness in analyzing visual media
-Perform and use sentiment and topic analysis on extracted visual content
-View potential differences/similarities between text-focused and visual media focused climate change posts

Additional Resources

Launch Presentation: