Scansoft, Inc.

Eric Fieleke

Joel M Gould

Michael L Elkins

Paul G Bamberg

Charles E Ingold

Kwangil Hwang

Kenneth J Bayse

Roger L Matus

Building a language model for use in speech recognition includes identifying without user interaction a source of text related to a user. Text is retrieved from the identified source of text and a language model related to the user is built from the retrieved text.

Language model for use in speech recognition

A computer is used to perform recorded actions. The computer receives recorded spoken utterances of actions. The computer then performs speech recognition on the recorded spoken utterances to generate texts of the actions. The computer then parses the texts to determine properties of the actions. After parsing the texts, permits the user to indicate that the user has reviewed one or more actions. The computer then automatically carries out the actions indicated as having been reviewed by the user.

Performing actions identified in recognized speech

Growing community of inventors

Medford, MA, United States of America

Eric Fieleke