Please Note: This project is not currently active. The content on this site is provided for reference and is not actively maintained.
  • A before-and-after snapshot of mood expressed about weather when a massive tornado hit Joplin Missouri. Dark orange colors indicate negative mood. Dark green represents positive mood. For those tweeting about the weather, mood shifted decidedly the day after the tornado struck.

  • Pulse is a set of analytical tools we are developing to track the ebb and flow of online conversations geographically. We extract and filter relevant content, add in rules for interpreting the expressed sentiment, and then pass jobs off to crowd-based workers to make judgments of sentiment. Pulse will enable us to identify key opportunities for introducing multimedia content into the dialogue.

Training the Cloud with the Crowd: Training A Google Prediction API Model Using CrowdFlower’s Workforce

by February 29, 2012

Want to see how people feel about the weather in San Diego, right now? Check out the weather sentiment prediction web app. Read about how Google Prediction API and Pulse data was used to train the cloud! More »

Tracking the Mood About Gas Prices on Twitter: A Case Study

by January 25, 2012

Have you ever wondered how the mood about gas prices fluctuates? Here we report results from a case study that reveals how this mood varies, or not, across the U.S. over a month-long period, including the Memorial Day weekend 2011. More »

Capturing Mood About Daily Weather From Twitter Posts

by September 29, 2011

Interest in the day's weather is widespread, with spirits dropping for many on gray, rainy days, and spirits soaring on sunny days. But, weather mood can be nuanced, with moods darkening when too much sun brings drought or a trace of snow causes a commuting mess in cities not accustomed to snow. Conversely, moods can turn upbeat when rain finally arrives in a parched area. Lots of people make comments about the weather on Twitter, though teasing out their mood is non trivial and currently requires an innovative approach based in crowdsourcing. This is the first comprehensive study of people's mood about the weather across the U.S. More »

More pulse blog posts