TERI RESEARCH INC  Providing the Information Advantage...                     

SERVICES       DEFENSE APPLICATIONS      KIOSK-COMMERCIAL       AMERICAN DISABILITIES ACT     

The Application Domain Interpreter (ADI) accepts textual strings from the Automatic Speech Recognizer (ASR) that may be a composite of word(s), phrases or sentences. ADI processing of these text strings is interpreted to determine the user intent and to pass actions to the Dialog Manager (DM). Speech Application Programming Interface (SAPI) protocol standards are used to request and pass text strings between the ASR and the ADI. Text string requests are routinely sent to the ASR after each ADI receipt of a text string.

User intent is determined by matching the domain focus with the valid transitional text strings identified in the Domain Database. Ambiguous or incomplete knowledge contained in the ASR text string is parsed by the ADI to determine the domain relevant information that is needed to complete the present intent focus. Alternatively, the ADI can determine if there is a shift in user focus that may be domain or rhetoric language based. A distinction is made by the ADI regarding any grammar ambiguity, to specify that either domain or rhetorical grammar is at issue.

The ADI uses knowledge-based logic to determine whether a valid text string has been received and what action is subsequently sent to the DM. Interpretation of the achievement of user intent, or partial information relevant to the Domain focus, is used to synchronize the NLS interface with the application. ADI completed action tokens or requests for additional information to complete the current focus are passed to the DM.

Natural language dictation transcription, with spoken command editing, is now practical with the advances made in automated profile learning and context adaptive technology.