Third International Conference on Spoken Language Processing (ICSLP 94)
We investigate devices which learn to understand and act upon spoken input. The ultimate goal of these speech understanding systems is to extract meaning from the speech signal. A crucial issue in engineering such devices is to quantify the information content of spoken natural language, and to be able to measure a machine's success in extracting that information. This paper discusses some recent progress in this direction.
Bibliographic reference. Gorin, Allen L. (1994): "Semantic associations, acoustic metrics and adaptive language acquisition", In ICSLP-1994, 79-82.