Systems and method for automatically configuring machine learning models

Abstract

Systems and methods for intelligently training a machine learning model includes: configuring a machine learning (ML) training data request for a pre-existing machine learning classification model; transmitting the machine learning training data request to each of a plurality of external training data sources, wherein each of the plurality of external training data sources is different; collecting and storing the machine learning training data from each of the plurality of external training data sources; processing the collected machine learning training data using a predefined training data processing algorithm; and in response to processing the collected machine learning training data, deploying a subset of the collected machine learning training data into a live machine learning model.

What is claimed: 
     
         1 . A system for rapidly training machine learning models of a machine learning-based conversational service, the system comprising:
 a machine learning configuration console that handles configuring machine learning models associated with the machine learning-based conversational service, wherein via the machine learning configuration console, the machine learning-based conversation service is configure to:
 [i] identify a seeding group that comprises a plurality of training data seeding samples; 
 [ii] generate a training data request based on the seeding group, wherein the training data request comprises the seeding group; 
 [iii] transmit the training data request to one or more third-party training data sources; 
 [iv] in response to the training data request, collecting raw machine learning training data from each of the one or more third-party training data sources; 
 [v] generate a fit score for each training data instance of the raw machine learning training data; 
 [vi] rank each of the training data instances of the raw machine learning training data based the fit score associated with each training data instance; 
 [vii] extract one or more training data instances, from the raw machine learning training data, having a fit score that does not satisfy a pruning threshold; and 
 [viii] in response to the extraction, train one or more machine learning classification models of the machine learning-based service using the raw machine learning training data that satisfy the pruning threshold. 
   
     
     
         2 . The system according to  claim 1 , wherein:
 the fit score indicates how well each training data instance from the raw machine learning data matches a text or a meaning of one or more of the plurality of the training data samples within a seeding group.   
     
     
         3 . The system according to  claim 1 , wherein:
 the pruning threshold comprises a minimum required fit score value for training data instances.   
     
     
         4 . The system according to  claim 1 , wherein:
 one or more subsets of the raw machine learning training data is generated at each of the one or more third-party training data sources based on the seeding group; and   each respective training data instance within the one or more subsets of the raw machine learning training data is provided a classification label by the one or more third-party training data sources that generated the respective training data instance.   
     
     
         5 . The system according to  claim 1 , wherein
 in response to detecting that a classification accuracy level of the one or more machine learning classification models that does not satisfy a predetermined threshold, automatically generating a notification requiring an update for improving the classification accuracy of the one or more machine learning classification models; and   generating the training data request based on the notification.   
     
     
         6 . The system according to  claim 1 , wherein
 prior to training the one or more machine learning classification models, simulating a performance of each of the one or more machine learning classification models using the raw machine learning training data that remains after the extraction; and   training the one or more machine learning classification models of the machine learning-based service using the raw machine learning training data based on the simulated performance.   
     
     
         7 . The system according to  claim 1 , wherein:
 the transmitting the machine learning training data request includes:
 identifying an input template for each of the one or more third-party training data sources, wherein the input template comprises input fields for receiving parameters for generating the machine learning training data at each of the one or more third-party training data sources, wherein the input template for each of the one or more third-party training data sources is different; 
 converting input data of the machine learning training data request to template input for the input template for each of the one or more third-party training data sources; and 
 feeding a respective input template having the converted input data of the machine learning training data request to a respective one of the one or more third-party training data sources. 
   
     
     
         8 . The system according to  claim 1 , wherein
 the one or more third-party training data sources comprise a plurality of third-party training data sources;   in response to the transmission of the training data request to each of the plurality of third-party training data sources, collecting in parallel the raw machine learning training from each of the plurality of third-party training data sources.   
     
     
         9 . The system according to  claim 1 , wherein:
 at least one of the one or more machine learning classification models comprises a competency classification machine learning model,   wherein the competency classification machine learning model is configured to generate a plurality of distinct competency classification labels,   each of the plurality of distinct competency classification labels corresponds to one competency of a plurality of areas of competencies of an artificially intelligent virtual assistant, and   a competency relates to a subject area of comprehension or aptitude of the artificially intelligent conversational system for which the artificially intelligent conversational system can interact with or provide a response to user input data.   
     
     
         10 . The system according to  claim 9 , wherein:
 the competency classification machine learning model comprises a single competency classification deep machine learning algorithm that is trained to detect each of the plurality of distinct competency classification labels, and   generating the competency classification label for the user input data includes selecting the competency classification label having a highest probability of matching an intent of the user input data.   
     
     
         11 . The system according to  claim 9 , wherein:
 the competency classification machine learning model comprises an ensemble of competency classification deep machine learning algorithms, wherein each competency classification deep machine learning algorithm of the ensemble is trained to detect a distinct competency classification label of the plurality of distinct competency classification labels, and   generating the competency classification label for the user input data includes selecting the competency classification label having a highest probability of matching an intent of the user input query.   
     
     
         12 . A method for rapidly training machine learning models of a machine learning-based conversational service, the method comprising:
 implementing a machine learning configuration console that handles configuring machine learning models associated with the machine learning-based conversational service;   identify, by the machine learning-based conversation service, a seeding group that comprises a plurality of training data seeding samples;   generate a training data request based on the seeding group, wherein the training data request comprises the seeding group;   transmit, via the machine learning configuration console, the training data request to one or more third-party training data sources;   in response to the training data request, collecting raw machine learning training data from each of the one or more third-party training data sources;   generating a fit score for each training data instance of the raw machine learning training data;   ranking each of the training data instances of the raw machine learning training data based the fit score associated with each training data instance;   extracting one or more training data instances, from the raw machine learning training data, having a fit score that does not satisfy a pruning threshold; and   in response to the extraction, training one or more machine learning classification models of the machine learning-based service using the raw machine learning training data that satisfy the pruning threshold.   
     
     
         13 . The method according to  claim 12 , wherein:
 the fit score indicates how well each training data instance from the raw machine learning data matches a text or a meaning of one or more of the plurality of the training data samples within a seeding group.   
     
     
         14 . The method according to  claim 12 , wherein:
 the pruning threshold comprises a minimum required fit score value for training data instances.   
     
     
         15 . The method according to  claim 12 , wherein:
 one or more subsets of the raw machine learning training data is generated at each of the one or more third-party training data sources based on the seeding group; and   each respective training data instance within the one or more subsets of the raw machine learning training data is provided a classification label by the one or more third-party training data sources that generated the respective training data instance.   
     
     
         16 . The method according to  claim 12 , wherein
 in response to detecting that a classification accuracy level of the one or more machine learning classification models that does not satisfy a predetermined threshold, automatically generating a notification requiring an update for improving the classification accuracy of the one or more machine learning classification models; and   generating the training data request based on the notification.   
     
     
         17 . The method according to  claim 12 , wherein
 prior to training the one or more machine learning classification models, simulating a performance of each of the one or more machine learning classification models using the raw machine learning training data that remains after the extraction; and   training the one or more machine learning classification models of the machine learning-based service using the raw machine learning training data based on the simulated performance.   
     
     
         18 . The method according to  claim 12 , further comprising:
 prior to the transmitting the training data request, identifying an input template for each of the one or more third-party training data sources; and   reformatting the training data request for the identified input template for each of the one or more third-party training data sources.