Methods and systems for automatic analysis of conversations between customer care agents and customers

Patent metadata

Field	Value
Publication number	US-9645994-B2
Application number	US-201414564170-A
Country	US
Kind code	B2
Filing date	Dec 9, 2014
Priority date	Dec 9, 2014
Publication date	May 9, 2017
Grant date	May 9, 2017

Abstract

The technical solution under the present disclosure automatically analyzes conversations between users by receiving a training dataset having a text sequence including sentences of a conversation between the users; extracting feature(s) from the training dataset based on features; providing equation(s) for a plurality of tasks, the equation(s) being a mathematical function for calculating value of a parameter for each of the tasks based on the extracted feature; determining value of the parameter for tasks by processing the equation(s); assigning label(s) to each of the sentences based on the determined value of the parameter, a first label being selected from a plurality of first labels, and a second label being selected from a number of second labels; and storing and maintaining with the database a pre-defined value of the parameter, first labels, conversations, second labels, a test dataset, equation(s), and pre-defined features.

First claim

What is claimed is:

1. A method for analyzing a natural language text sequence of a conversation exchanged between multiple users of a platform, the method comprising:
extracting, by one or more processors, at least one feature from a training dataset that includes a natural language text sequence of a conversation exchanged between multiple users of a platform based on a pre-defined features stored at a database, wherein the natural language text sequence comprises a plurality of sentences, each of the plurality of sentences including a number of words;
defining, by the one or more processors, at least one equation for two or more tasks to be completed;
calculating, by the one or more processors, a parameter value for each of the two or more tasks using the equation, the parameter value being based on the at least one feature extracted from the training dataset;
selecting, by the one or more processors, one or more first labels from a plurality of first labels and one or more second labels from a plurality of second labels based on the calculated parameter value and applying the selected one or more first and second labels to each of the plurality of sentences of the natural language sequence of the training dataset; and
based on the selected one or more first and second labels, indicating to the multiple users of the platform, by the one or more processors, that the two or more tasks are complete.

2. The method of claim 1 , wherein the at least one feature is extracted based on the plurality of pre-defined features, the plurality of pre-defined features including at least one of a ‘Word 1-grams and 2-grams’, ‘Segment position in thread’, ‘Segment position in tweet’, ‘Sender’, ‘Contains email’, ‘#Upper case’, ‘#Punctuation’, ‘#Special punctuation’, ‘Positive Sentiment’, ‘Negative Sentiment’, ‘Category of previous segment’, ‘Category of previous tweet/author’, and ‘Category of previous tweet’.

3. The method of claim 2 , wherein the two or more tasks include at least one of a task to determine a status of an issue in the conversation exchanged between multiple users of the platform, and a task to determine a nature of the conversation exchanged between multiple users of the platform.

4. The method of claim 3 , wherein calculating the parameter value includes processing, by the one or more processors, the at least one equation for the two or more tasks for a pre-defined number of iterations.

5. The method of claim 4 , wherein the plurality of first labels comprises at least one of ‘Open’, ‘Solved’, ‘Closed’, and ‘Changed channel’.

6. The method of claim 5 , wherein the plurality of first labels are associated with the at least one task of determining determine a status of an issue.

7. The method of claim 6 , wherein the plurality of second labels comprises at least one of a ‘Complaint’, ‘Apology’, ‘Answer’, ‘Receipt’, ‘Compliment’, ‘Response to positivity’, ‘Request’, ‘Greeting’, ‘Thank’, ‘Announcement’, ‘Solved’, and ‘Other’.

8. The method of claim 7 , wherein the plurality of second labels are associated with the at least one task of determining the nature of the conversation.

9. The method of claim 8 , wherein the at least one equation includes a common parameter for each of the two or more tasks to be completed.

10. The method of claim further comprising:
comparing, by one or more processors, the calculated parameter value for each of the two or more tasks to a pre-defined parameter value stored at the database, the pre-defined parameter value being calculated based on an evaluation of a test dataset, wherein the test dataset includes a pre-defined set of sentences that each include at least one word.

11. The method of claim 10 , wherein the natural language text sequence of the training dataset includes a natural language text sequence of a conversation exchanged between a customer care agent and multiple customers of a platform, wherein the natural language text sequence of the conversation exchanged between the customer care agent and multiple customers of a platform comprises a plurality of sentences exchanged between the customer care agent and multiple customers using a social media platform.

12. A system for automated analysis of a natural language text sequence of a conversation exchanged between multiple users of a platform, the system comprising:
one or more processors configured to:
store a plurality of pre-defined features, two or more tasks to be completed, a plurality of first labels, and a plurality of second labels;
extract at least one feature from a training dataset based on the plurality of pre-defined features stored at the database, wherein the training dataset includes a natural language text sequence of a conversation exchanged between multiple users of a platform, the natural language text sequence including a plurality of sentences;
define at least one equation for the two or more tasks to be completed;
calculate a parameter value for each of the two or more tasks using the equation, the parameter value being calculated based on the at least one extracted feature and stored at the database;
select one or more first labels from the plurality of stored first labels and one or more second labels from the plurality of stored second labels based on the calculated parameter value;
apply the selected one or more first and second labels to each of the plurality of sentences of the natural language text sequence of the training dataset; and
indicate to the multiple users of the platform when the two or more tasks are complete, the indication being based on the applied first and second labels.

13. The system of claim 12 , wherein the at least one feature is extracted based on the plurality of pre-defined features, the plurality of pre-defined features including at least one of a ‘Word 1-grams and 2-grams’, ‘Segment position in thread’, ‘Segment position in tweet’, ‘Sender’, ‘Contains email’, ‘#Upper case’, ‘#Punctuation’, ‘#Special punctuation’, ‘Positive Sentiment’, ‘Negative Sentiment’, ‘Category of previous segment’, ‘Category of previous tweet/author’, and ‘Category of previous tweet’.

14. The system of claim 13 , wherein the two or more tasks include at least one of a task to determine a status of an issue in the conversation exchanged between multiple users of the platform, and a task to determine a nature of the conversation exchanged between multiple users of the platform.

15. The system of claim 14 , wherein calculating the parameter value further includes:
processing, by the one or more processors, the at least one equation for the two or more tasks for a pre-defined number of iterations.

16. The system of claim 15 , wherein the plurality of first labels comprises at least one of ‘Open’, ‘Solved’, and ‘Changed channel’.

17. The of claim 16 , wherein the plurality of first labels are associated with the at least one task of determining a status of an issue.

18. The system of claim 17 , wherein the plurality of second labels comprises at least one of a ‘Complaint’, ‘Apology’, ‘Answer’, ‘Receipt’, ‘Compliment’, ‘Response to positivity’, ‘Request’, ‘Greeting’, ‘Thank’, ‘Announcement’, ‘Solved’, and ‘Other’.

19. The system of claim 18 , wherein the plurality of second labels are associated with the at least one task of determining the nature of the conversation.

20. The system of claim 19 , wherein the at least one equation includes a common parameter for each of the two or more tasks to be completed.

21. The system of claim 20 , wherein the one or more processors are

Assignees

• Xerox Corp
• Conduent Business Services Llc

Inventors

• Agarwal Arvind
• Kataria Saurabh
• Sun Tong
• Bhatia Sumit

Classifications

• G06Q10/40

  Business processes related to social networking or social networking services · CPC title

• G06Q30/016 Primary

  After-sales · CPC title

• G06F40/35

  Discourse or dialogue representation · CPC title

• G06Q30/00

  Commerce · CPC title

• G06F40/289

  Phrasal analysis, e.g. finite state techniques or chunking · CPC title