Systems and methods for automatically categorizing unstructured data and improving a machine learning-based dialogue system

Abstract

Systems and methods for building a response for a machine learning-based dialogue agent includes implementing machine learning classifiers that predict slot segments of the utterance data based on an input of the utterance data; predict a slot classification label for each of the slot segments of the utterance data; computing a semantic vector value for each of the slot segments of the utterance data; assessing the semantic vector value of the slot segments of the utterance data against a multi-dimensional vector space of structured categories of dialogue, wherein the assessment includes: for each of a distinct structured categories of dialogue computing a similarity metric value; selecting one structured category of dialogue from the distinct structured categories of dialogue based on the computed similarity metric value for each of distinct structured categories; and producing a response to the utterance data.

What is claimed: 
     
         1 . A method for building a response for a machine learning-based dialogue agent based on mapping unstructured data of an utterance to one of a plurality of distinct categories of dialogue, the method comprising:
 identifying utterance data;   implementing one or more machine learning classifiers that generate one or more predictions for handling the utterance data;   computing a semantic vector value for one or more slot segments of the utterance data;   assessing the semantic vector value for each of the one or more slot segments of the utterance data against a multi-dimensional vector space of structured categories of dialogue;   wherein:
 the multi-dimensional vector space comprises vector values for a plurality of distinct structured categories of dialogue, each of the plurality of distinct structured categories of dialogue having an expanded hierarchical structure, 
   wherein the assessment includes:
 (i) computing a first similarity metric value for each pairing of the one or more slot segments of the utterance data and each distinct top-level category of the expanded hierarchical structure based on the evaluation; 
 (ii) computing a second similarity metric value for each pairing of the one or more slot segments of the utterance data and each of one or more distinct sub-categories of dialogue of a given expanded hierarchical structure; 
 (iii) computing an average similarity metric value for each the plurality of distinct structured categories of dialogue based on a sum of the first similarity metric value for each distinct top-level category and the second similarity metric value of the given expanded hierarchical structure; 
   selecting one structured category of dialogue from the plurality of distinct structured categories of dialogue based on the computed average similarity metric value for each of the plurality of distinct structured categories of dialogue of the multi-dimensional vector space; and   producing a response to the utterance data that is communicated via the machine learning-based dialogue agent based at least on the selected one structured category of dialogue.   
     
     
         2 . A system for mapping unstructured data of an utterance to one of a plurality of distinct categories, the system comprising:
 a machine learning-based automated dialogue service implemented by one or more hardware computing servers that:
 identify utterance data; 
 implement one or more machine learning classifiers that generate one or more predictions for handling the utterance data; 
 compute a semantic vector value for one or more slot segments of the utterance data; 
 assess the semantic vector value for each of the one or more slot segments of the utterance data against a multi-dimensional vector space of structured categories of dialogue; 
 wherein:
 the multi-dimensional vector space comprises vector values for a plurality of distinct structured categories of dialogue, each of the plurality of distinct structured categories of dialogue having an expanded hierarchical structure, 
 
 wherein the assessment includes:
 (i) computing a first similarity metric value for each pairing of the one or more slot segments of the utterance data and each distinct top-level category of the expanded hierarchical structure based on the evaluation; 
 (ii) computing a second similarity metric value for each pairing of the one or more slot segments of the utterance data and each of one or more distinct sub-categories of dialogue of a given expanded hierarchical structure; 
 (iii) computing an average similarity metric value for each the plurality of distinct structured categories of dialogue based on a sum of the first similarity metric value for each distinct top-level category and the second similarity metric value of the given expanded hierarchical structure; 
 
 select one structured category of dialogue from the plurality of distinct structured categories of dialogue based on the computed average similarity metric value for each of the plurality of distinct structured categories of dialogue of the multi-dimensional vector space; and 
 produce a response to the utterance data that is communicated via the machine learning-based dialogue agent based at least on the selected one structured category of dialogue.