Computer-implemented system and method for transcription error reduction
A computer-implemented system and method for transcription error reduction is provided. A transcribed value is assigned to each utterance obtained from a user during a call and a confidence score is assigned to each transcribed value. An accuracy threshold is applied to the confidence scores and the transcribed values that satisfy the accuracy threshold are incorporated in a message. A grouping is generated for at least one of the utterances associated with one such transcribed value that fails to satisfy the accuracy threshold. The grouping includes the at least one utterance and related utterances from other calls. Further transcribed values for at least a portion of the utterances in the grouping are received from human transcribers. The remaining utterances in the grouping are provided to the human transcribers when the further transcribed values differ. The further transcribed value for the at least one utterance is incorporated in the message.
1. A computer-implemented system for transcription error reduction, comprising:
a processor to execute modules, comprising:
an assignment module configured to assign a transcribed value to each utterance obtained from a user during a call and a confidence score for each transcribed value;
an application module configured to apply an accuracy threshold to the confidence scores;
an incorporation module configured to incorporate the transcribed values that satisfy the accuracy threshold in a message;
a grouping module configured to generate a grouping for at least one of the utterances associated with one such transcribed value that fails to satisfy the accuracy threshold, the grouping comprising the at least one utterance and related utterances from other calls;
a receipt module configured to receive from one or more human transcribers a further transcribed value for each utterance in a portion of the grouping;
a delivery module configured to provide any remaining utterances in the grouping to the one or more human transcribers when the further transcribed values differ; and
a placement module configured to incorporate the further transcribed value for the at least one utterance in the message.
2. A system according to claim 1 , further comprising:
an identification module configured to identify the related utterances based on one or more of a similarity of characteristics of the call and the other calls and a range of similarity shared between the transcribed value for the at least one utterance and other transcribed values for the utterances from the other calls.
3. A system according to claim 2 , further comprising:
a selection module configured to select a predetermined number of the related utterances within a predetermined time from identifying the at least one utterance associated with one such transcribed value that fails to satisfy the accuracy threshold.
4. A system according to claim 3 , further comprising:
a time determination module configured to determine the predetermined time based on at least one of an accuracy of transcription required and a type of the utterances received from the user comprising one of recorded utterances and live utterances.
5. A system according to claim 1 , further comprising:
a display module configured to place in the message the transcribed value for the at least one utterance prior to incorporating the further transcribed value in the message and to display the transcribed value differently from the other transcribed values in the message that satisfy the accuracy threshold.
6. A system according to claim 5 , further comprising:
a replacement module configured to replace the transcribed value for the at least one utterance with an identification value for the at least one utterance and the confidence score for the at least one utterance.
7. A system according to claim 6 , further comprising:
a further replacement module configured to replace in the message the identification value for the at least one utterance with the further transcribed value.
8. A system according to claim 1 , further comprising at least one of:
a database to store the further transcribed value and the at least one utterance with a grammar; and
a grammar generator configured to generate a new grammar with the further transcribed value and the at least one utterance.
9. A system according to claim 1 , further comprising:
a grammar selection module configured to select the assigned transcribed values for the utterances obtained from the user from one or more grammars, wherein at least one of the grammars is selected based on a characteristic of the user.
10. A system according to claim 1 , wherein the utterances obtained from the user are received during a live call with an agent.
11. A computer-implemented method for transcription error reduction, comprising:
assigning a transcribed value to each utterance obtained from a user during a call and a confidence score for each transcribed value;
applying an accuracy threshold to the confidence scores;
incorporating the transcribed values that satisfy the accuracy threshold in a message;
generating a grouping for at least one of the utterances associated with one such transcribed value that fails to satisfy the accuracy threshold, the grouping comprising the at least one utterance and related utterances from other calls;
receiving from one or more human transcribers a further transcribed value for each utterance in a portion of the grouping;
providing any remaining utterances in the grouping to the one or more human transcribers when the further transcribed values differ; and
incorporating the further transcribed value for the at least one utterance in the message.
12. A method according to claim 11 , further comprising:
identifying the related utterances based on one or more of a similarity of characteristics of the call and the other calls and a range of similarity shared between the transcribed value for the at least one utterance and other transcribed values for the utterances from the other calls.
13. A method according to claim 12 , further comprising:
selecting a predetermined number of the related utterances within a predetermined time from identifying the at least one utterance associated with one such transcribed value that fails to satisfy the accuracy threshold.
14. A method according to claim 13 , further comprising:
determining the predetermined time based on at least one of an accuracy of transcription required and a type of the utterances received from the user comprising one of recorded utterances and live utterances.
15. A method according to claim 11 , further comprising:
prior to incorporating the further transcribed value in the message for the at least one utterance, placing the transcribed value for the at least one utterance in the message and displaying the transcribed value differently from the other transcribed values in the message that satisfy the accuracy threshold.
16. A method according to claim 15 , further comprising:
replacing the transcribed value for the at least one utterance with an identification value for the at least one utterance and the confidence score for the at least one utterance.
17. A method according to claim 16 , further comprising:
replacing in the message the identification value for the at least one utterance with the further transcribed value.
18. A method according to claim 11 , further comprising at least one of:
storing the further transcribed value and the at least one utterance with a grammar; and
generating a new grammar with the further transcribed value and the at least one utterance.
19. A method according to claim 11 , further comprising:
selecting the assigned transcribed values for the utterances obtained from the user from one or more grammars, wherein at least one of the grammars is selected based on a characteristic of the user.
20. A method according to claim 11 , wherein the utterances obtained from the user are received during a live call with an agent.