Systems and methods for monitoring speech data labelers
View Patent ↗Systems and methods herein use an annotation guide to label utterances and speech data with a call type. A system practicing the method embodiment monitors labelers of speech data by presenting via a processor a test utterance to a labeler, receiving input from the labeler that selects a particular call type from a list of call types and determining via the processor if the labeler labeled the test utterance correctly. Based on the determining step, the system revises the annotation guide, retrains the labeler, and/or alters the test utterance.
1. In a system where an annotation guide is used to label utterances in speech data with a call type, a method for monitoring a labeler of the speech data, the method comprising:
presenting, via a processor of a computing device, a test utterance to the labeler;
generating a determination indicating whether the labeler correctly labeled the test utterance as one call type from a list of call types; and
based on the determination, performing at least one of revising the annotation guide, retraining the labeler, and altering the test utterance.
2. The method of claim 1 , wherein the test utterance is selected from a group of labeled utterances.
3. The method of claim 2 , wherein the group of labeled utterances is generated based on the annotation guide.
4. The method of claim 1 , wherein generating the determination further comprises receiving input from the labeler that selects an additional call type from the list of call types for the test utterance.
5. The method of claim 1 , wherein the test utterance has an existing call type.
6. The method of claim 1 , wherein if the test utterance is altered, the method further comprises presenting a second test utterance related to a different call type.
7. The method of claim 1 , wherein if the test utterance is altered, the method further comprises presenting a second test utterance that tests for more utterances within a same call type.
8. A system for monitoring a labeler of the speech data, the system comprising:
a processor;
a memory storing instructions for controlling the processor to perform a method comprising:
presenting a test utterance to the labeler;
generating a determination indicating whether the labeler correctly labeled the test utterance as one call type from a list of call types; and
performing at least one of revising the annotation guide, retraining the labeler, and altering the test utterance based on the determination.
9. The system of claim 8 , the instructions further comprising selecting the test utterance from a group of labeled utterances.
10. The system of claim 9 , wherein the group of labeled utterances is generated using the annotation guide.
11. The system of claim 8 , the instructions further comprising receiving input from a labeler that selects an additional call type from the list of call types for the test utterance.
12. The system of claim 8 , wherein the test utterance has an existing call type.
13. The system of claim 8 , wherein if the test utterance is altered, the instructions further comprise presenting another test utterance related to a different call type.
14. The system of claim 8 , wherein if the test utterance is altered, the instructions further comprise presenting another test utterance that tests for more utterances within the same call type.
15. A non-transitory computer-readable storage medium storing instructions for controlling a computing device to monitor a labeler of the speech data where an annotation guide is used to label utterances in the speech data with a call type, the instructions comprising:
presenting a test utterance to the labeler;
generating a determination indicating whether the labeler correctly labeled the test utterance as one call type from a list of call types; and
based on the determination, performing at least one of revising the annotation guide, retraining the labeler, and altering the test utterance.
16. The non-transitory computer-readable storage medium of claim 15 , wherein if the test utterance is altered, the instructions further comprise presenting another test utterance related to a different call type.
17. The non-transitory computer-readable storage medium of claim 15 , wherein presenting the test utterance to the labeler further comprises selecting the test utterance from a group of labeled utterances.
18. The non-transitory computer-readable storage medium of claim 17 , wherein selecting the test utterance from a group of labeled utterances further comprises generating the group of labeled utterances using the annotation guide.
19. The non-transitory computer-readable storage medium of claim 15 , further comprising receiving input from the labeler that selects an additional call type from the list of call types for the test utterance.
20. The non-transitory computer-readable storage medium of claim 15 , wherein the test utterance has an existing call type.