Automatic extension of an ATN knowledge base

A computer program is described that acquires much of its knowledge from conversations among operators on Morse code radio networks. The system consists of a learning component and a language understander. The learning component extends a ‘core’ augmented transition network (ATN) knowledge base by generalizing from sentences taken from scripts of actual conversations. The extensions enable the understanding component to process a large number of sentences that are syntactically and semantically similar to the examples.

View this article in the ACM Digital Library.