a news portal that tracks references to people, places, & things appearing in news text from over 1000 domestic & international news sources every day, & analyzes meaningful relationships between them. it uses natural language processing techniques to track entity references in news sources, as well as their the temporal & geographical distribution over time.

the resulting information is presented by an extensive collection of data visualizations, including a relational network (e.g. showing connections between people), a news sector distribution (e.g. sports, entertainment, news, business), a sentiment analysis graph (e.g emotional state in various communities), & a locative world & US popularity maps (of which some can be animated over time or appear as "battle maps" comparing the localized reporting of 2 news entities).

[link: textmap.com & press.jrc.it|via datamining.typepad.com]