BBN Technologies To Teach Machines To Read

BBN Technologies has been awarded $29.7 million in funding by the Defense Advanced Research Projects Agency (DARPA) under the Machine Reading program in a contract awarded by the Air Force Research Laboratory (AFRL).

The goal of the Machine Reading program is to develop a revolutionary, automated reading system that bridges the gap between naturally occurring text and the artificial intelligence reasoning systems that need such knowledge.

Although this intelligent learning system would initially be used to automate military intelligence analysis, it could also enable a variety of civilian applications. For example, as more and more of the world’s libraries are converted to digital text, this system could provide unprecedented access and automated analysis, allowing for vastly expanded cultural awareness and historical and cultural research.

Under this contract, BBN will leverage its expertise in natural language processing and distillation to develop a universal text engine that captures knowledge from text and transforms it into the formal representations required by artificial intelligence systems.

A central goal of the research effort is to develop techniques that can generalize across the linguistic structure and content of diverse documents to extract relations and axioms directly from text rather than relying on a knowledge engineer to encode such information.

A related goal is to develop techniques capable of performing such automatic extraction of text at the massive scale available on the World Wide Web. Over the course of the five-year program, BBN’s system will be tested against increasingly complex targets, including its ability to learn axioms from text and to read and digest vast quantities of Web text.