# US 8,751,232 B2 · System and method for targeted tuning of a speech recognition system

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 8,751,232 B2
- Document ID: US8751232B2
- Status: Granted
- Kind: B2
- Application number: 13/760,897
- Filed: 2013-02-06
- Published: 2013-06-13
- Granted: 2014-06-10
- Earliest claimed: 2004-08-12
- Assignees: AT&T Intellectual Property I, L.P.
- Applicants: AT&T Intellectual Property I, L.P.
- Inventors: Robert R. Bushey; Benjamin Anthony Knott; John Mills Martin
- Web page: https://patentgrep.com/patent/US8751232B2

## CPC

- G10L 15/19: Grammatical context, e.g. disambiguation of the recognition hypotheses based on word sequence rules
- G10L 15/063: {Training}
- G10L 15/065: Adaptation
- G10L 15/22: Procedures used during a speech recognition process, e.g. man-machine dialogue
- G10L: SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING

## Abstract

A system and method of targeted tuning of a speech recognition system are disclosed. A particular method includes detecting that a frequency of occurrence of a particular type of utterance satisfies a threshold. The method further includes tuning a speech recognition system with respect to the particular type of utterance.

## Claims

### Claim 1

A method comprising:
detecting that a frequency of occurrence of a particular type of utterance satisfies a threshold; and
in response to detecting that the frequency satisfies the threshold, tuning a speech recognition system with respect to the particular type of utterance.

### Claim 2 (depends on 1)

The method of claim 1, further comprising determining the frequency based on a group of received utterances.

### Claim 3 (depends on 1)

The method of claim 1, wherein the threshold is determined by a system administrator.

### Claim 4 (depends on 1)

The method of claim 1, wherein the threshold is user programmable.

### Claim 5 (depends on 1)

The method of claim 1, wherein tuning the speech recognition system includes inputting a collection of utterances of the particular type of utterance into a learning module of the speech recognition system.

### Claim 6 (depends on 5)

The method of claim 5, wherein inputting the collection of utterances includes playing one or more files that represent recordings of the particular type of utterance.

### Claim 7 (depends on 1)

The method of claim 1, wherein system recognition of the particular type of utterance is dependent on a particular speaker.

### Claim 8 (depends on 1)

The method of claim 1, wherein system recognition of the particular type of utterance is independent of a particular speaker.

### Claim 9 (depends on 1)

The method of claim 1, wherein the utterance is one of a single word spoken by a speaker, a phrase spoken by the speaker, or a sentence spoken by the speaker.

### Claim 10 (depends on 1)

The method of claim 1, wherein the utterance corresponds to a request that indicates an action to be taken on an object.

### Claim 11 (depends on 10)

The method of claim 10, wherein the request is one of a request to pay a bill, a request for an account balance, a request to change services, a request to acquire services, a request to cancel services, a request to schedule a payment, or a request to reconnect services.

### Claim 12

A speech recognition system, comprising:
a processor; and
a tuning engine comprising instructions that are executable by the processor to tune a speech module with respect to a particular type of utterance in response to the processor determining that a frequency of occurrence of the particular type of utterance satisfies a threshold.

### Claim 13 (depends on 12)

The speech recognition system of claim 12, further comprising a targeting engine including instructions that are executable by the processor to:
determine the frequency of occurrence of the particular type of utterance; and
determine that the frequency of occurrence of the particular type of utterance satisfies the threshold.

### Claim 14 (depends on 12)

The speech recognition system of claim 12, wherein the threshold is from a table of thresholds, the thresholds corresponding to a plurality of utterance types.

### Claim 15 (depends on 12)

The speech recognition system of claim 12, further comprising an accuracy engine including instructions executable by the processor to determine an accuracy of recognition of the particular type of utterance.

### Claim 16 (depends on 15)

The speech recognition system of claim 15, wherein the accuracy of recognition of the particular type of utterance is compared to an accuracy threshold.

### Claim 17 (depends on 16)

The speech recognition system of claim 16, wherein the tuning engine is responsive to the accuracy engine.

### Claim 18 (depends on 15)

The speech recognition system of claim 15, wherein the frequency of occurrence of the particular type of utterance depends on the accuracy of recognition of the particular type of utterance.

### Claim 19

A computer readable storage device comprising instructions that, when executed by a processor, cause the processor to perform operations comprising:
detecting that a frequency of occurrence of a particular type of utterance satisfies a threshold; and
in response to detecting that the frequency satisfies the threshold, tuning a speech recognition system with respect to the particular type of utterance.

### Claim 20 (depends on 19)

The computer readable storage device of claim 19, wherein the operations further include:
calculating a hit rate associated with the particular type of utterance;
calculating a false alarm value associated with the particular type of utterance; and
maintaining a record of performance of the speech recognition system.