Adaptive Language Acquisition For An Airline Information Subsystem

The goal of this paper is to extend our principles and mechanisms for adaptive language acquisition to moderate complexity tasks. In our previous work, a basic principle was that the primary function of language is to convey meaning, with the consequence that language acquisition involves gaining the capability of decoding that meaning. This led us to investigate a language acquisition mechanism based on connectionist methods, in which the network builds associations between messages and meaningful responses to them. As a task increases in complexity, however, so does the mapping from message to meaning.