Automated Media Cloud Pipelines for Tracking AI Data Center News Coverage

This AIMES Lab research poster presents an automated Media Cloud pipeline for analyzing how news coverage of a topic differs by region and outlet type. Applied to 9,919 articles on AI data centers from national outlets and local outlets in Virginia, Texas, and Ohio, the study uses LDA topic modeling and LLM-assisted analysis to reveal a sharp divide: national coverage frames data centers as a technology and growth story, while local coverage frames them as a resource and land-use story.

The research was conducted by Anastasiia Boltkova and Adam Perhala, in collaboration with AIMES Lab Director Prof. John P. Wihbey and Dan Zedek. It details the data collection, topic modeling, and source attribution methods behind the pipeline, contributing to ongoing research in computational journalism, media analysis, and AI coverage.