System and method for providing words or phrases to be uttered by members of a crowd and processing the utterances in crowd-sourced campaigns to facilitate speech analysis

Abstract

Systems and methods of providing text related to utterances, and gathering voice data in response to the text are provide herein. In various implementations, an identification token that identifies a first file for a voice data collection campaign, and a second file for a session script may be received from a natural language processing training device. The first file and the second file may be used to configure the mobile application to display a sequence of screens, each of the sequence of screens containing text of at least one utterance specified in the voice data collection campaign. Voice data may be received from the natural language processing training device in response to user interaction with the text of the at least one utterance. The voice data and the text may be stored in a transcription library.

What is claimed is: 
     
         1 . A computer-implemented method of providing crowd-sourced campaigns to facilitate speech analysis, the method being implemented in a computer system having one or more physical processors programmed with computer program instructions that, when executed by the one or more physical processors, program the computer system to perform the method, the method comprising:
 receiving, by the computer system, from a natural language processing training device, an identification token containing a first portion and a second portion, the first portion identifying a first file associated with a voice data collection campaign, and the second portion identifying a second file associated with a session script, the session script comprising one or more prompts to be provided at a mobile application on the natural language processing training device;   identifying, by the computer system, one or more audits based on the identification token;   conducting, by the computer system, the identified one of more audits based on the identification token via the mobile application;   receiving, by the computer system, a plurality of audit responses from the mobile application;   determining, by the computer system, a number of failed audit responses based on the received plurality of audit responses;   storing, by the computer system, the number of failed audits in association with a device identifier that identifies the natural language processing training device;   determining, by the computer system, whether the number of failed audits for the natural language processing training device exceeds a failed audits threshold; and   responsive to a determination that the number of failed audits exceeds a failed audits threshold, barring, by the computer system, the natural language processing training device from the voice data collection campaign.   
     
     
         2 . The method of  claim 1 , wherein conducting the one or more audits comprises providing gold standard questions, captions that are not machine-readable, and/or audio that is not understandable to machines. 
     
     
         3 . The method of  claim 1 , the method further comprising:
 gathering, by the computer system, a first filename corresponding to the first file;   gathering, by the computer system, a second filename corresponding to the second file; and   creating, by the computer system, the identification token using the first filename and the second filename.   
     
     
         4 . The method of  claim 1 , wherein one or more of the first file and the second file comprises a JavaScript Object Notation (JSON) file. 
     
     
         5 . The method of  claim 1 , wherein the session script identifies one or more prompts, and wherein each prompt comprises text of at least one utterance, the method further comprising:
 providing, by the computer system, the one or more prompts to a user via the mobile application on the natural language processing training device;   receiving, by the computer system, voice data from the natural language processing training device in response to user interaction with the prompt; and   storing, by the computer system, the voice data and the prompt in a transcription library.   
     
     
         6 . The method of  claim 5 , wherein the at least one utterance comprises one or more of a syllable, a word, a phrase, or a variant thereof. 
     
     
         7 . The method of  claim 5 , wherein the user interaction comprises a selection of a touch-screen button instructing the mobile application to record the voice data. 
     
     
         8 . The method of  claim 5 , the method further comprising:
 obtaining, by the computer system, a decibel level of the received voice data;   obtaining, by the computer system, campaign data based on the first file, wherein the campaign data includes a specification of a minimum decibel level;   determining, by the computer system, whether the decibel level for the received voice data exceeds the specified minimum decibel level;   responsive to a determination that the decibel level exceeds the specified minimum decibel level, storing, by the computer system, the voice data and the text of the at least one utterance in the transcription library; and   responsive to a determination that the decibel level does not exceed the specified minimum decibel level, not storing, by the computer system, the voice data and the text of the at least one utterance in the transcription library.   
     
     
         9 . The method of  claim 5 , the method further comprising:
 obtaining, by the computer system, a decibel level of the received voice data;   obtaining, by the computer system, campaign data based on the first file, wherein the campaign data includes a specification of a maximum decibel level;   determining, by the computer system, whether the decibel level for the received voice data exceeds the specified maximum decibel level;   responsive to a determination that the decibel level does not exceed the specified maximum decibel level, storing, by the computer system, the voice data and the text of the at least one utterance in the transcription library; and   responsive to a determination that the decibel level exceeds the specified maximum decibel level, not storing, by the computer system, the voice data and the text of the at least one utterance in the transcription library.   
     
     
         10 . The method of  claim 5 , the method further comprising:
 obtaining, by the computer system, an audio duration of the received voice data;   obtaining, by the computer system, campaign data based on the first file, wherein the campaign data includes a specification of a maximum audio duration;   determining, by the computer system, whether the audio duration exceeds the specified maximum audio duration;   responsive to a determination that the audio duration does not exceed the specified maximum audio duration, storing, by the computer system, the voice data and the text of the at least one utterance in the transcription library; and   responsive to a determination that the audio duration exceeds the specified maximum audio duration, not storing, by the computer system, the voice data and the text of the at least one utterance in the transcription library.   
     
     
         11 . The method of  claim 5 , the method further comprising:
 receiving, by the computer system, a request to repeat the one or more prompts to a second user via the mobile application on the natural language processing training device;   obtaining, by the computer system, campaign data based on the first file, wherein the campaign data specifies whether the natural language processing training device can repeat the one or more prompts; and   determining, by the computer system, whether the natural language processing training device can repeat the one or more prompts based on the campaign data.   
     
     
         12 . The method of  claim 5 , the method further comprising:
 obtaining, by the computer system, campaign data based on the first file, wherein the campaign data specifies whether a calibration screen should be displayed via the mobile application on the natural language processing training device; and   instructing, by the computer system, the natural language processing training device to display the calibration screen via the mobile application based on the campaign data.   
     
     
         13 . The method of  claim 1 , wherein the natural language processing training device comprises one or more of a mobile phone, a tablet computing device, a laptop, and a desktop. 
     
     
         14 . The method of  claim 1 , wherein the voice data collection campaign is configured to collect demographic information related to a user of the natural language processing training device. 
     
     
         15 . A system for providing crowd-sourced campaigns to facilitate speech analysis, the system comprising:
 one or more physical processors programmed with computer program instructions that, when executed by the one or more physical processors, program the one or more physical processors to:
 receive from a natural language processing training device an identification token containing a first portion and a second portion, the first portion identifying a first file associated with a voice data collection campaign, and the second portion identifying a second file associated with a session script, the session script comprising one or more prompts to be provided at a mobile application on the natural language processing training device; 
 identify one or more audits based on the identification token; 
 conduct the identified one of more audits based on the identification token via the mobile application; 
 receive a plurality of audit responses from the mobile application; 
 determine a number of failed audit responses based on the received plurality of audit responses; 
 store the number of failed audits in association with a device identifier that identifies the natural language processing training device; 
 determine whether the number of failed audits for the natural language processing training device exceeds a failed audits threshold; and 
 responsive to a determination that the number of failed audits exceeds a failed audits threshold, bar the natural language processing training device from the voice data collection campaign. 
   
     
     
         16 . The system of  claim 15 , wherein to conduct the one or more audits, the one or more processors are further programmed to:
 provide gold standard questions, captions that are not machine-readable, and/or audio that is not understandable to machines.   
     
     
         17 . The system of  claim 15 , wherein the one or more processors are further programmed to:
 gather a first filename corresponding to the first file;   gather a second filename corresponding to the second file; and   create the identification token using the first filename and the second filename.   
     
     
         18 . The system of  claim 15 , wherein one or more of the first file and the second file comprises a JavaScript Object Notation (JSON) file. 
     
     
         19 . The system of  claim 15 , wherein the session script identifies one or more prompts, each prompt comprising text of at least one utterance, wherein the one or more processors are further programmed to:
 provide the one or more prompts to a user via the mobile application on the natural language processing training device;   receive voice data from the natural language processing training device in response to user interaction with the prompt; and   store the voice data and the prompt in a transcription library.   
     
     
         20 . The system of  claim 19 , wherein the at least one utterance comprises one or more of a syllable, a word, a phrase, or a variant thereof. 
     
     
         21 . The system of  claim 19 , wherein the user interaction comprises a selection of a touch-screen button instructing the mobile application to record the voice data. 
     
     
         22 . The system of  claim 19 , wherein the one or more processors are further programmed to:
 obtain a decibel level of the received voice data;   obtain campaign data based on the first file, wherein the campaign data includes a specification of a minimum decibel level;   determine whether the decibel level for the received voice data exceeds the specified minimum decibel level;   responsive to a determination that the decibel level exceeds the specified minimum decibel level, store the voice data and the text of the at least one utterance in the transcription library; and   responsive to a determination that the decibel level does not exceed the specified minimum decibel level, not store the voice data and the text of the at least one utterance in the transcription library.   
     
     
         23 . The system of  claim 19 , wherein the one or more processors are further programmed to:
 obtain a decibel level of the received voice data;   obtain campaign data based on the first file, wherein the campaign data includes a specification of a maximum decibel level;   determine whether the decibel level for the received voice data exceeds the specified maximum decibel level;   responsive to a determination that the decibel level does not exceed the specified maximum decibel level, store the voice data and the text of the at least one utterance in the transcription library; and   responsive to a determination that the decibel level exceeds the specified minimum decibel level, not store the voice data and the text of the at least one utterance in the transcription library.   
     
     
         24 . The system of  claim 19 , wherein the one or more processors are further programmed to:
 obtain an audio duration of the received voice data;   obtain campaign data based on the first file, wherein the campaign data includes a specification of a maximum audio duration;   determine whether the audio length exceeds the specified maximum audio duration;   responsive to a determination that the audio duration does not exceed the specified maximum audio duration, store the voice data and the text of the at least one utterance in the transcription library; and   responsive to a determination that the audio duration exceeds the specified maximum audio duration, not store the voice data and the text of the at least one utterance in the transcription library.   
     
     
         25 . The system of  claim 19 , wherein the one or more processors are further programmed to:
 receive a request to repeat the one or more prompts to a second user via the mobile application on the natural language processing training device;   obtain campaign data based on the first file, wherein the campaign data specifies whether the natural language processing training device can repeat the one or more prompts; and   determine whether the natural language processing training device can repeat the one or more prompts based on the campaign data.   
     
     
         26 . The system of  claim 19 , wherein the one or more processors are further programmed to:
 obtain campaign data based on the first file, wherein the campaign data specifies whether a calibration screen should be displayed via the mobile application on the natural language processing training device; and   instruct the natural language processing training device to display the calibration screen via the mobile application based on the campaign data.   
     
     
         27 . The system of  claim 15 , wherein the natural language processing training device comprises one or more of a mobile phone, a tablet computing device, a laptop, and a desktop. 
     
     
         28 . The system of  claim 15 , wherein the voice data collection campaign is configured to collect demographic information related to a user of the natural language processing training device.