Method for processing the output of a speech recognizer
First Claim
1. A method for processing speech, comprising:
- providing an automated speech processing system having a command processor and at least one memory;
receiving a natural language speech input generated by a microphone;
semantically parsing the received natural language speech input using an automated statistical processor with respect to a plurality of predetermined command grammars in the automated speech processing system, the plurality of predetermined grammars defining mutually inconsistent command outcomes from the command processor, said semantically parsing selectively excluding predetermined command grammars having respective command outcomes inconsistent with the previously received natural language speech;
determining;
if the semantic parsing of the received natural language speech input corresponds to a single non-excluded command outcome according to a respective command grammar, and is complete for reliable processing by the command processor, then processing the command with the command processor, according to the single non-excluded command grammar and exiting said determining;
if the received natural language speech input corresponds to a plurality of command outcomes, or is not complete for reliable processing according to a plurality of non-excluded predetermined command grammars, then;
prompting a user for further natural language speech input dependent on at least one of the plurality of non-excluded predetermined command grammars, the prompting comprising feedback representing an identification of at least one command type putatively recognized, and information required to reduce correspondence to a plurality of non-excluded predetermined command grammars or to increase completeness, in dependence on a relationship of the previously received natural language speech input and at least one command grammar of the plurality of non-excluded predetermined command grammars,reparsing the further natural language speech input with the automated statistical processor in conjunction with previously parsed natural language speech input, anditerating said determining; and
if an abort, fail or cancel condition is detected in the natural language speech input, exiting said determining.
1 Assignment
0 Petitions
Accused Products
Abstract
A method for processing speech, comprising semantically parsing a received natural language speech input with respect to a plurality of predetermined command grammars in an automated speech processing system; determining if the parsed speech input unambiguously corresponds to a command and is sufficiently complete for reliable processing, then processing the command; if the speech input ambiguously corresponds to a single command or is not sufficiently complete for reliable processing, then prompting a user for further speech input to reduce ambiguity or increase completeness, in dependence on a relationship of previously received speech input and at least one command grammar of the plurality of predetermined command grammars, reparsing the further speech input in conjunction with previously parsed speech input, and iterating as necessary. The system also monitors abort, fail or cancel conditions in the speech input.
133 Citations
20 Claims
-
1. A method for processing speech, comprising:
-
providing an automated speech processing system having a command processor and at least one memory; receiving a natural language speech input generated by a microphone; semantically parsing the received natural language speech input using an automated statistical processor with respect to a plurality of predetermined command grammars in the automated speech processing system, the plurality of predetermined grammars defining mutually inconsistent command outcomes from the command processor, said semantically parsing selectively excluding predetermined command grammars having respective command outcomes inconsistent with the previously received natural language speech; determining; if the semantic parsing of the received natural language speech input corresponds to a single non-excluded command outcome according to a respective command grammar, and is complete for reliable processing by the command processor, then processing the command with the command processor, according to the single non-excluded command grammar and exiting said determining; if the received natural language speech input corresponds to a plurality of command outcomes, or is not complete for reliable processing according to a plurality of non-excluded predetermined command grammars, then; prompting a user for further natural language speech input dependent on at least one of the plurality of non-excluded predetermined command grammars, the prompting comprising feedback representing an identification of at least one command type putatively recognized, and information required to reduce correspondence to a plurality of non-excluded predetermined command grammars or to increase completeness, in dependence on a relationship of the previously received natural language speech input and at least one command grammar of the plurality of non-excluded predetermined command grammars, reparsing the further natural language speech input with the automated statistical processor in conjunction with previously parsed natural language speech input, and iterating said determining; and if an abort, fail or cancel condition is detected in the natural language speech input, exiting said determining. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10)
-
-
11. A system for processing speech, comprising:
-
a port configured to receive a natural language speech input from a user through a microphone; a memory configured to store the received natural language speech input; an automated statistical speech processor configured to semantically parse the natural language speech input with respect to a plurality of predetermined command grammars defining mutually inconsistent command outcomes; an automated speech processor configured to; exclude predetermined command grammars of the plurality of predetermined grammars having respective command outcomes inconsistent with the previously received natural language speech; determine; if the semantically parsed natural language speech input corresponds to a single non-excluded command grammar and is complete for reliable processing, then storing the command grammar for execution and exiting the determining; if the previously received natural language speech input corresponds to a plurality of non-excluded predetermined command grammars or is not complete for reliable processing, then; prompting a user for further natural language speech input to further exclude predetermined command grammars based on inconsistency with the previously received natural language speech input or increase completeness with respect to reliable processing, present the further natural language speech input to the automated statistical speech processor for reparsing in conjunction with previously parsed natural language speech input, and iterate the determination; and if an abort, fail or cancel condition is detected in the natural language speech input, then exiting the determination said determining. - View Dependent Claims (12, 13, 14, 15, 16)
-
-
17. A method for processing speech, comprising:
-
receiving a speech input from a user through a microphone, and a context of the received speech input; testing an output of a speech recognizer to determine; whether it probabilistically indicates that a command to be processed is present, if the command to be processed is probabilistically present, parsing the command with respect to a plurality of predetermined command grammars representing a plurality of alternate command grammars consistent with the received speech input, and whether the probabilistically present command to be processed is unambiguous and complete subject to the context with high statistical reliability; in dependence on the testing, either; processing the probabilistically present command if unambiguous and complete subject to the context with high statistical reliability or prompting the user to provide specific inputs to reduce the ambiguity or increase completeness subject to the context of the probabilistically present command, to increase the statistical reliability of the probabilistically present command; and determining whether the user seeks to abort processing of an incomplete or ambiguous command, and in dependence on said determining, halting processing of the probabilistically present command. - View Dependent Claims (18, 19, 20)
-
Specification