When
Join us for the College of Information Science Colloquium Series, featuring Laura W. Dozal.
Attend in-person or via Zoom (registration is required):
Friday, January 23, 2026
Presentation 12-1 p.m.
Grad Workshop 1-2 p.m. (in-person only)
An AI/MLOps pipeline runs synthetically generated image data through a multimodal large language model to generate labels and captions for each image. The labels and captions are semantically evaluated and then integrated into a network structure to examine thematic representations and how they are grouped. The synthetically generated images focus on a specific case study involving a social movement. The process uses quantitative analysis and human-in-the-loop evaluation to identify patterns within the network structure and summarize the overall narrative found in the image collection.