Processor for Open source INTelligence (POINT)

Knexient proposes POINT, a processing system for OSI on the Web. Specifically, POINT's text processing engine finds the primary or main event in online new stories and relates the event to its location and time of occurrence. The location of the event is converted into Google's KML language and presented as a text overlay on GoogleEarth, enabling military image analysts to quickly aggregate relevant information about events in particular locations. POINT uses an innovative hierarchical classification system, the Hierarchical Document Classifier (HDC), to classify primary events down to the granularity of particular actions (e.g., a suicide bombing), and to classify very large volumes of semantically heterogenous text, such as (to take a few examples) reports about meetings, bombings, elections, or vehicle accidents. Specifically, HDC classifies news according to the specific event that is primary in the story (e.g., a suicide bombing), and the event context for the specific event (the IraqWar). The event pair combine to tell the analyst what the story is about based on the events described.