twXray technology provides a Twitter user with a profile of what he or she is talking about on Twitter. Given a Twitter username, the system retrieves the most recent few hundred tweets from that user, and analyzes them topically using a statistical categorizer. After assigning each tweet to a topical category, the system presents a graphical view of the percentage of the tweets that concern each category. The system also displays some of the specific tweets for each category.
Because the technology is statistical in nature, it will sometimes make mistakes in categorizing a given tweet. Overall, though, the system provides a good picture of what you are thinking about and talking about on Twitter.
twXray was originally developed as part of a larger system, Twitter Profiling, that provides users with recommendations about news, blogs, and other media that might be interesting to them, based on their Twitter activity.
Original development team:Shawn O'Banion Product development:Jennifer Wilson |
Design:Jeremy Gilbert Design assistance:Rich Gordon |
twXray (pronounced "twix-ray") was developed at Northwestern University's Knight News Innovation Lab, with support from the John S. and James L. Knight Foundation and the National Science Foundation. If you have any comments or suggestions, email us at twxray@gmail.com.