IARPA wants to learn about new methods for extracting and coding ‘societal events’

The Intelligence Community wants to find better ways to extract and code what are called “societal events” — particularly social, political, epidemiological, cyber, economic, counterintelligence and technology events — from both structured and unstructured databases.

To that end, the Intelligence Advanced Research Projects Activity (IARPA) issued a Request for Information (RFI) on February 5, which seeks detailed information from potential suppliers of such data extracting methodologies. Responses are due to IARPA by March 21, according to the agency’s special notice.

The Intelligence Community is eager to be able to forecast a broad set of well-defined social events that are relevant to national security. “For many of the events of interest, the ground truth is developed by extracting events from unstructured data, often news text,” explains IARPA’s notice.

IARPA wants to learn more about “event coding” for special classes of societal events, as well as training an experienced event coder to code new event classes.

The agency wants to hear about new methodologies and how they compare with existing event coders, such as SERIF or TABARI. “A discussion of how these methods perform across different data sources (e.g., news report versus blogs) is of interest,” says IARPA, “as are approaches to deduplicating event entries.”

The notice asks interested parties to submit their responses (including a half-page executive summary and a maximum of five pages of description) by March 21.

Among several questions the notice poses is the following: “What structured databases exist that contain both historical and frequently updated events relevant to national security?”

Further information about this RFI is available from Jason Matheny, of IARPA, at dni-iarpa-rfi-14-04@iarpa.gov.