Computer-implemented system and method for efficiently reducing transcription error during a call
View Patent ↗A computer-implemented system and method for efficiently reducing transcription error during a call is provided. A group of similar questionable utterances is monitored. Each questionable utterance is assigned a transcribed value and is selected from a different call. A sample of the similar questionable utterances is selected from the group and provided to a human transcriber. A further transcribed value for each of the similar questionable utterances in the sample is received from the human transcriber. The further transcribed values for the similar questionable utterances in the sample are compared. Each of the remaining similar questionable utterances in the group are provided to the human transcriber when the compared further transcribed values of the similar questionable utterances in the sample are different.
1. A computer-implemented system for efficiently reducing transcription error during a call, comprising:
at least one processor;
a group of similar questionable utterances, wherein each questionable utterance is assigned a transcribed value and is selected from a different call;
a sample module to select a sample of the similar questionable utterances from the group and to provide the sample to a human transcriber;
a transcription module to receive from the human transcriber a further transcribed value for each of the similar questionable utterances in the sample;
a comparison module to compare the further transcribed values for the similar questionable utterances in the sample; and
a delivery module to provide to the human transcriber each of the remaining similar questionable utterances in the group when the compared further transcribed values of the similar questionable utterances in the sample are different.
2. A system according to claim 1 , further comprising:
a group generation module to generate the group of similar questionable utterances by identifying a questionable utterance, comparing other questionable utterances to the identified questionable utterance, and designating as the group of similar questionable utterances, the identified questionable utterance and the other questionable utterances each with a similarity to the identified questionable utterance that satisfies a predetermined threshold.
3. A system according to claim 2 , further comprising:
an identification module to identify the other questionable utterances with similarities that satisfy the predetermined threshold within a predetermined amount of time.
4. A system according to claim 3 , further comprising:
a determination module to determine the predetermined amount of time based on at least one of whether the call is in real-time and a desired transcription accuracy.
5. A system according to claim 2 , further comprising:
a similarity module to determine the similarity for each of the other questionable utterances based on at least one of initial transcribed values, confidence scores, a range of similarity, and call characteristics.
6. A system according to claim 1 , further comprising:
an assignment module to assign the further transcribed value to one such similar questionable utterance in the sample when the transcribed value for that similar questionable utterance differs from the further transcribed value.
7. A system according to claim 1 , further comprising:
a confirmation module to confirm the transcribed value assigned to one such similar questionable utterance in the sample when the further transcribed value is the same as the assigned transcribed value.
8. A system according to claim 1 , further comprising:
a transcription module to assign the transcribed value to each similar questionable utterance by recording a voice stream from each different call, parsing the voice stream from each different call into utterances, and selecting the transcribed value for each utterances based on a grammar.
9. A system according to claim 8 , further comprising:
a message generation module to combine the transcribed values for one such voice stream and to generate a transcribed message for that voice stream from the combined transcribed values.
10. A system according to claim 8 , further comprising:
an identification module to identify at least one of the utterances for one such voice stream as the questionable utterances, comprising:
an assignment module to assign a confidence score to each of the transcribed values for that voice stream; and
a threshold module to identify one or more of the utterances in the voice stream that have transcribed values with confidence scores that fall below a predetermined threshold as the questionable utterances.
11. A computer-implemented method for efficiently reducing transcription error during a call, comprising:
monitoring a group of similar questionable utterances, wherein each questionable utterance is assigned a transcribed value and is selected from a different call;
selecting a sample of the similar questionable utterances from the group and providing the sample to a human transcriber;
receiving from the human transcriber a further transcribed value for each of the similar questionable utterances in the sample;
comparing the further transcribed values for the similar questionable utterances in the sample; and
providing to the human transcriber each of the remaining similar questionable utterances in the group when the compared further transcribed values of the similar questionable utterances in the sample are different.
12. A method according to claim 11 , further comprising:
generating the group of similar questionable utterances, comprising:
identifying a questionable utterance;
comparing other questionable utterances to the identified questionable utterance; and
designating as the group of similar questionable utterances, the identified questionable utterance and the other questionable utterances each with a similarity to the identified questionable utterance that satisfies a predetermined threshold.
13. A method according to claim 12 , further comprising:
identifying the other questionable utterances with similarities that satisfy a predetermined threshold within the predetermined amount of time.
14. A method according to claim 13 , further comprising:
determining the predetermined amount of time based on at least one of whether the call is in real-time and a desired transcription accuracy.
15. A method according to claim 12 , further comprising:
determining the similarity for each of the other questionable utterances based on at least one of initial transcribed values, confidence scores, a range of similarity, and call characteristics.
16. A method according to claim 11 , further comprising:
assigning the further transcribed value to one such similar questionable utterance in the sample when the transcribed value for that similar questionable utterance differs from the further transcribed value.
17. A method according to claim 11 , further comprising:
confirming the transcribed value assigned to one such similar questionable utterance in the sample when the further transcribed value is the same as the assigned transcribed value.
18. A method according to claim 11 , further comprising:
assigning the transcribed value to each similar questionable utterance, comprising:
recording a voice stream from each different call;
parsing the voice stream from each different call into utterances; and
selecting the transcribed value for each utterances based on a grammar.
19. A method according to claim 18 , further comprising:
combining the transcribed values for one such voice stream; and
generating a transcribed message for that voice stream from the combined transcribed values.
20. A method according to claim 18 , further comprising:
identifying at least one of the utterances for one such voice stream as the questionable utterances, comprising:
assigning a confidence score to each of the transcribed values for that voice stream; and
identifying one or more of the utterances in the voice stream that have transcribed values with confidence scores that fall below a predetermined threshold as the questionable utterances.