Automatically responding to a request of a user

• US 9,715,496 B1
• Filed: 09/01/2016
• Issued: 07/25/2017
• Est. Priority Date: 07/08/2016
• Status: Active Grant

• Current Assignee: Asapp, Inc.
• Original Assignee: Asapp, Inc.
• Inventors: Sapoznik, Gustavo, Dai, Hui, Hackman, Joseph, Henry, Shawn
• Primary Examiner(s): Shah, Paras D
• Application Number:
  US15/254,008
• Time in Patent Office: 327 Days
• Field of Search: 704270, 704275, 379 8801
• CPC Class Codes: G06F 16/2237 Vectors, bitmaps or matrices
• CPC Class Codes: G06F 16/245 Query processing
• CPC Class Codes: G06F 16/3329 Natural language query formulation or dialogue systems
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06F 16/9024 Graphs; Linked lists G06F16/9027 takes precedence
• CPC Class Codes: G06F 16/90332 Natural language query formulation or dialogue systems
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 40/205 Parsing
• CPC Class Codes: G06F 40/211 Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
• CPC Class Codes: G06F 40/216 using statistical methods
• CPC Class Codes: G06F 40/35 Discourse or dialogue representation
• CPC Class Codes: G06F 40/40 Processing or translation of natural language natural language analysis G06F40/20; semantic analysis G06F40/30
• CPC Class Codes: G06N 20/10 using kernel methods, e.g. support vector machines [SVM]
• CPC Class Codes: G06N 3/04 Architecture, e.g. interconnection topology
• CPC Class Codes: G06N 3/044 Recurrent networks, e.g. Hopfield networks
• CPC Class Codes: G06N 3/045 Combinations of networks
• CPC Class Codes: G06N 3/047 Probabilistic or stochastic networks
• CPC Class Codes: G06N 3/08 Learning methods
• CPC Class Codes: G06N 3/088 Non-supervised learning, e.g. competitive learning
• CPC Class Codes: G06N 5/01 Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
• CPC Class Codes: G06N 5/022 : Knowledge engineering; Knowledge acquisition
• CPC Class Codes: G06Q 10/06311 : Scheduling, planning or task assignment for a person or group
• CPC Class Codes: G06Q 10/1053 : Employment or hiring
• CPC Class Codes: G06Q 10/107 : Computer-aided management of electronic mailing [e-mailing]
• CPC Class Codes: G06Q 30/01 : Customer relationship services
• CPC Class Codes: G06Q 30/016 : After-sales
• CPC Class Codes: G10L 15/26 : Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: H04L 41/04 : Network management architectures or arrangements
• CPC Class Codes: H04L 51/02 : using automatic reactions or user delegation, e.g. automatic replies or chatbot-generated messages
• CPC Class Codes: H04L 51/52 : for supporting social networking services
• CPC Class Codes: H04L 67/02 : based on web technology, e.g. hypertext transfer protocol [HTTP]
• CPC Class Codes: H04L 67/306 : User profiles
• CPC Class Codes: H04L 67/53 : using third party service providers
• CPC Class Codes: H04M 3/42382 : Text-based messaging services in telephone networks such as PSTN/ISDN, e.g. User-to-User Signalling or Short Message Service for fixed networks
• CPC Class Codes: H04M 3/5141 : Details of processing calls and other types of contacts in an unified manner unified messaging in packet-switching networks H04L51/56
• CPC Class Codes: H04M 3/5183 : Call or contact centers with computer-telephony arrangements

Abstract

A user may request assistance or information using natural language, such as by speaking or entering a text. A response may be automatically determined to the request of the user be performing semantic processing to understand the request and formulate an appropriate response. A response may be automatically determined, be computing features from the request, using the features to select a node of a graph corresponding to the request, and then causing an action to be performed to provide a response to the user. For example, the action may include providing information to the user, requesting further information from the user, or connecting the user with a customer service representative.

Claims

1. A computer-implemented method for automatically responding to a request of a customer, the method comprising:

receiving, at a server computer, text of a message corresponding to a customer request, wherein the message is received from a computing device of the customer;

computing, by a server computer, a plurality of features from text of the message;

obtaining, by a server computer, a graph comprising a plurality of nodes, wherein each node of the graph corresponds to a node selector classifier and an action, wherein each node selector classifier receives the plurality of features as inputs and processes the inputs to output scores, and wherein each node of the graph corresponds to a different type of a customer request;

obtaining, by a server computer, a first node selector classifier corresponding to a first node of the graph;

processing, by a server computer, the plurality of features with the first node selector classifier to generate a first plurality of scores comprising a score for traversing to one or more child nodes of the first node;

selecting, by a server computer, a second node of the graph using the first plurality of scores, wherein the second node is a child node of the first node;

obtaining, by a server computer, a second node selector classifier corresponding to the second node, wherein the second node selector classifier is different from the first node selector classifier;

processing, by a server computer, the plurality of features with the second node selector classifier to generate a second plurality of scores comprising (i) a score indicating a match between the customer request and the second node and (ii) a score for traversing to one or more child nodes of the second node;

selecting, by a server computer, the second node using the second plurality of scores; and

causing, by a server computer, an action corresponding to the selected second node to be performed, wherein the action comprises responding to the customer request by sending a computer-generated message to the customer or electronically connecting the customer with a customer service representative.

2. The computer-implemented method of claim 1, wherein the score indicating a match between the customer request and the second node is a largest score of the second plurality of scores.

3. The computer-implemented method of claim 1, wherein the first node selector classifier is independent of the second node selector classifier.

4. The computer-implemented method of claim 1, wherein receiving text of a message corresponding to a customer request comprises performing speech recognition on an audio signal comprising speech of the customer.

5. The computer-implemented method of claim 1, wherein the message is received via a text message, electronic mail, a web server, or an application running on the computing device of the customer.

6. The computer-implemented method of claim 1, wherein selecting the second node comprises using a greedy search algorithm.

7. The computer-implemented method of claim 1, wherein computing the first plurality of scores comprises using information about the customer.

8. The computer-implemented method of claim 1, wherein the first plurality of scores is not used in generating the second plurality of scores.

9. The computer-implemented method of claim 1, comprising:

receiving an identifier of the customer;

obtaining data about the customer from a data store comprising data about a plurality of customers; and

wherein the first classifier is configured to generate the first plurality of scores by processing the features and the data about the customer.

10. The computer-implemented method of claim 9, wherein the second classifier does not process the data about the customer.

11. The computer-implemented method of claim 9, wherein the data about the customer comprises a customer profile, a location of the customer, billing data, payment data, or services provided to the customer.

12. A system for automatically responding to a request of a customer, the system comprising:

at least one server computer comprising at least one processor and at least one memory, the at least one server computer configured to;

receive, at the at least one server computer, text of a message corresponding to a customer request, wherein the message is received from a computing device of the customer;

compute, by the at least one server computer, a plurality of features from text of the message;

obtain, by the at least one server computer, a graph comprising a plurality of nodes, wherein each node of the graph corresponds to a node selector classifier and an action, wherein each node selector classifier receives the plurality of features as inputs and processes the inputs to output scores, and wherein each node of the graph corresponds to a different type of a customer request;

obtain, by the at least one server computer, a first node selector classifier corresponding to a first node of the graph;

process, by the at least one server computer, the plurality of features with the first node selector classifier to generate a first plurality of scores comprising a score for traversing to one or more child nodes of the first node;

select, by the at least one server computer, a second node of the graph using the first plurality of scores, wherein the second node is a child node of the first node;

obtain, by the at least one server computer, a second node selector classifier corresponding to the second node, wherein the second node selector classifier is different from the first node selector classifier;

process, by the at least one server computer, the plurality of features with the second node selector classifier to generate a second plurality of scores comprising (i) a score indicating a match between the customer request and the second node and (ii) a score for traversing to one or more child nodes of the second node;

select, by the at least one server computer, the second node using the second plurality of scores; and

cause, by the at least one server computer, an action corresponding to the selected second node to be performed, wherein the action comprises responding to the customer request by sending a computer-generated message to the customer or electronically connecting the customer with a customer service representative.

13. The system of claim 12, wherein the second node is associated with a plurality of actions, and wherein the at least one server computer is configured to select an action from the plurality of actions.

14. The system of claim 13, wherein the at least one server computer is configured to select the action using information about the customer.

15. The system of claim 12, wherein the graph is a directed acyclic graph.

16. The system of claim 12, wherein the at least one server computer is configured to select the second node using a beam search algorithm.

17. The system of claim 12, wherein the at least one server computer is configured to compute the second plurality of scores using information.

18. The system of claim 12, wherein the first classifier comprises a logistic regression classifier.

19. One or more non-transitory computer-readable media comprising computer executable instructions that, when executed, cause at least one processor to perform actions comprising:

receiving, at a server computer, text of a message corresponding to a customer request, wherein the message is received from a computing device of the customer;

computing, by a server computer, a plurality of features from text of the message;

obtaining, by a server computer, a graph comprising a plurality of nodes, wherein each node of the graph corresponds to a node selector classifier and an action, wherein each node selector classifier receives the plurality of features as inputs and processes the inputs to output scores, and wherein each node of the graph corresponds to a different type of a customer request;

obtaining, by a server computer, a first node selector classifier corresponding to a first node of the graph;

processing, by a server computer, the plurality of features with the first node selector classifier to generate a first plurality of scores comprising a score for traversing to one or more child nodes of the first node;

selecting, by a server computer, a second node of the graph using the first plurality of scores, wherein the second node is a child node of the first node;

obtaining, by a server computer, a second node selector classifier corresponding to the second node, wherein the second node selector classifier is different from the first node selector classifier;

processing, by a server computer, the plurality of features with the second node selector classifier to generate a second plurality of scores comprising (i) a score indicating a match between the customer request and the second node and (ii) a score for traversing to one or more child nodes of the second node;

selecting, by a server computer, the second node using the second plurality of scores; and

causing, by a server computer, an action corresponding to the selected second node to be performed, wherein the action comprises responding to the customer request by sending a computer-generated message to the customer or electronically connecting the customer with a customer service representative.

20. The one or more non-transitory computer-readable media of claim 19, wherein the actions comprise traversing from a root node of the graph to the first node.

21. The one or more non-transitory computer-readable media of claim 19, wherein computing the plurality of features comprises encoding words of the text of the message and processing the encoded words with a neural network.

22. The one or more non-transitory computer-readable media of claim 19, wherein the first node selector classifier comprises a logistic regression classifier.

23. The one or more non-transitory computer-readable media of claim 19, wherein:

selecting the second node of the graph using the first plurality of scores comprises comparing at least one of the first plurality of scores to a first threshold; and

selecting the second node of the graph using the second plurality of scores comprises comparing at least one of the second plurality of scores to a second threshold, wherein the second threshold is not equal to the first threshold.

24. The one or more non-transitory computer-readable media of claim 19, wherein selecting the second node comprises using information about the customer.