| US 7,519,534 B2 | ||
| Speech controlled access to content on a presentation medium | ||
| Scott Lynn Maddux, San Francisco, Calif. (US); Adam Jordan, Oakland, Calif. (US); Jody Williams, San Carlos, Calif. (US); and Harry W. Printz, San Francisco, Calif. (US) | ||
| Assigned to AgileTV Corporation, Menlo Park, Calif. (US) | ||
| Filed on Oct. 30, 2003, as Appl. No. 10/699,543. | ||
| Claims priority of provisional application 60/483268, filed on Jun. 26, 2003. | ||
| Claims priority of provisional application 60/422561, filed on Oct. 31, 2002. | ||
| Prior Publication US 2004/0193426 A1, Sep. 30, 2004 | ||
| Int. Cl. G10L 15/00 (2006.01); G10L 15/22 (2006.01) | ||
| U.S. Cl. 704—255 [704/270; 704/275] | 7 Claims |

| 1. A method for linking grammars into a hierarchy of operations, comprising the steps of:
establishing various grammars each grammar including various utterances and, for each utterance, the following associated
attributes: (1) an indication of whether the utterance is explicitly chained to a further grammar, or (2) contextual information
indicating a type of data implicitly specified by the utterance;
where each one of the various grammars further includes, for each utterance that is explicitly chained to a further grammar,
a chained command attribute identifying the further grammar for activating responsive to a user issuing that utterance while
said further grammar is activated for speech recognition;
where the various grammars include command grammars and information-type grammars, and:
utterances in the command grammars form commands to control a manner of presenting video programs;
utterances in the information-type grammars form keywords pertaining to content of video programs;
accepting a statement comprising a series of uninterrupted user utterances containing more than one user command for controlling
a video program; and
performing a series of operations by activating a chained series of one or more grammars recognized from the statement, wherein
for each grammar that contains a user utterance that is explicitly chained to a further grammar, activating the further grammar;
and for each grammar that contains a user utterance having contextual information associated within, determining which options
to make available to user next, thus activating a further grammar based on the contextual information of a preceding grammar,
wherein the further grammar at least partially includes the options.
|