The rise of the social web marks a technology shift in the evolution of the web. But as interesting as the technology shift has been the cultural shift that has accompanied it, where activities that were traditionally considered private have increasingly become public. People who had previously written in a journal, kept family photo albums, and talked to their friends by phone or e-mail are now increasingly writing blogs, posting their photos on public photo sharing sites, and having conversations via public wall posts on social networks.
The result of this is that the web has become home to a mass archive of human communication and sentiment, containing large amounts of valuable information that are not well handled by traditional methods of retrieval and presentation. While the data is now available to answer queries like "how did the U.S. feel when Obama was elected?", search engines don't currently handle queries like this well.
Our research in this area involves extraction and visualization of sentiment data from the social web, and in building interfaces for real-time querying, analysis, and presentation of social and emotional data.
The key projects that we have done in this area are We Feel Fine, a system that crawls the world's blogs every minute, and extracts sentences that contain the words "I feel" or "I am feeling", and I Want You To Want Me, a system that continuously crawls online dating sites for expressions of self-definition and desire.