Case Study: University of Exeter uncovers media's impact on public behavior with Labelbox

University of Exeter uses ML to reveal how the media impacts human behavior

The University of Exeter researchers faced the challenge of building large, high-quality training datasets to study how news and social media shape public attitudes and behavior — including multi-label image classification (using CNNs and transfer learning) for COVID-19 news frames and U-Net segmentation for analyzing nature content in organizational social posts. Their existing tools fell short, so they adopted the Labelbox platform and the Labelbox Workforce to manage and scale labeling efforts.

Labelbox provided a managed labeling solution and on-demand workforce that let the team rapidly scale data collection and obtain consistently high-quality annotations, enabling reliable training of CNN and U-Net models. With Labelbox’s platform and workforce, the researchers created a large, rich dataset that supported analysis at a previously unattainable scale and allowed more accurate determination of media influence on individual behavior.

University of Exeter

Travis Coan

Researcher