Christopher M Danforth

Christopher McLaughlin Danforth is an American computer scientist and professor of applied mathematics at the University of Vermont. He is known for his work in computational social science, particularly for developing the Hedonometer, a tool for measuring collective mood and happiness using sentiment analysis of social media data.

Danforth co-directs the Computational Story Lab at the Vermont Complex Systems Institute with Peter Sheridan Dodds, where he leads research on large-scale sociotechnical systems, exploring human behavior through analysis of social media and other digital data sources.

Education and career Danforth earned his Ph.D. in applied mathematics, with early research focused on numerical weather prediction and applications of chaos theory to improve weather forecasts. He holds the position of Professor in the Department of Mathematics and Statistics at the University of Vermont and serves as Associate Director of the Vermont Complex Systems Institute.

Research

Hedonometer and sentiment analysis In 2007, Danforth collaborated with Peter Sheridan Dodds to develop the Hedonometer, an instrument designed to measure happiness and emotional states in large populations through analysis of social media text. The project involved surveying native speakers of multiple languages to rate words on a happiness scale, creating a weighted dictionary of emotional valence for approximately 10,000 words in ten languages.

The Hedonometer measures real-time emotional trends by analyzing text from Twitter and other social media platforms. Research using the tool has examined emotional responses to major events, geographical variations in happiness, and temporal patterns in collective mood. Studies have documented applications in analyzing events such as the Ferguson protests, where the tool identified significant shifts in word usage patterns reflecting community sentiment.

The study found that photographs posted by individuals with depression tended to be bluer, darker, and grayer in color compared to those posted by non-depressed individuals. The computational model achieved a 70% accuracy rate in identifying depression, outperforming the average 42% diagnostic success rate of general practitioners in unassisted depression diagnosis. Notably, the model could detect signs of depression in photographs posted before individuals received clinical diagnoses, suggesting potential for early screening applications.

Computational Story Lab Danforth co-directs the Computational Story Lab with Peter Sheridan Dodds, a research group focused on large-scale, system-level problems across sociology, nonlinear dynamics, networks, ecology, and physics. The lab's research combines observational data, mathematical models, and computational simulations to understand complex phenomena.

The lab has developed several tools and datasets for the research community, including: Hedonometer: A public website providing real-time happiness measurements from Twitter data StoryWrangler: A platform for analyzing phrase popularity across 150 billion tweets

Funding and support Danforth's research has been supported by multiple organizations, including the National Science Foundation, the National Institutes of Health, the James S. McDonnell Foundation, the Office of Naval Research, NASA, and MITRE Corporation.

In 2023, as part of a group of researchers in the Vermont Complex Systems Institute, he helped procure a National Science Foundation award to study the science of stories. This 5-year grant was called The Science of Online Corpora, Knowledge, and Stories (SOCKS).

In 2025, Danforth led a team that received a $2.1 million grant from the National Science Foundation to create IceCore, a high-performance computing cluster for artificial intelligence research. The system is designed to support approximately 1,000 researchers across Vermont and New England studying topics including infectious diseases, computational social science, and large language models. He has developed video lecture series on nonlinear dynamics that are freely available online. His most highly cited works include research on temporal patterns of happiness in social networks, the universal positivity bias in human language, and computational methods for mental health screening.

Selected publications