DARPA puts faith in text analytics

Faced with an avalanche of text information and the need to keep Americans safe, DARPA has announced plans to put machine learning, advanced computer processes able to make sense of large amounts of unstructured data, to use.

"We want the ability to mitigate ambiguity in text by stripping away filters that can cloud meaning and by rejecting false information. To be successful, the technology needs to look beyond what is explicitly expressed in text to infer what is actually meant," said DARPA's Bonnie Dorr.

The overall goal of DARPA's project is largely the same as companies hoping to use big data in business – to increase the quality of decisions. Current methods of reviewing intelligence data can be too slow to produce helpful results for military decision-makers. A recent estimate found that up to 90 percent of collected data can't be reviewed in time to be useful.

The government as a whole has become committed to the idea of big data analytics as a problem solving technology. The overall federal commitment to unstructured data research totals $200 million. In addition to DARPA and the Department of Defense, the National Institute of Health and National Science Foundation are slated to receive funds.