Systems and methods for facilitating dialogue mining

Patent metadata

Field	Value
Publication number	US-10635695-B2
Application number	US-201715587183-A
Country	US
Kind code	B2
Filing date	May 4, 2017
Priority date	Dec 23, 2013
Publication date	Apr 28, 2020
Grant date	Apr 28, 2020

Abstract

The disclosure is related to mining of text to derive information from the text that is useful for a variety of purposes. The text mining process can be implemented in a service oriented industry such as a call center, where a customer and an agent engage in a dialog, e.g., to discuss product/service related issues. The messages in dialogues between the customers and the agents are tagged with features that describe an aspect of the conversation. The text mining process can mine various dialogues and identify a set of features and messages based on prediction algorithms. The identified set of features and messages can be used to infer an intent of a particular customer for contacting the agent, and to generate a recommendation based on the determined intent.

First claim

The invention claimed is:

1. A method, comprising:
analyzing, by a computing system, a plurality of messages exchanged between a customer and an agent to extract a plurality of aspects of a dialog between the customer and the agent from one or more of the plurality of messages;
tagging, by the computing system, a message of the plurality of messages that comprise the dialog with a set of features to generate a set of tagged features;
generating, by the computing system, a storage object for the message, the storage object including the message and the set of tagged features that are based upon the plurality of aspects of the dialog which are extracted from the message; and
storing, by the computing system, the storage object in a storage system, the storage system including a plurality of storage objects that correspond to a second plurality of messages exchanged between a plurality of customers and a plurality of agents;
wherein at least some of the second plurality of messages are tagged with one or more features depicting a set of aspects extracted from a specified dialog between one of the plurality of customers and one of the plurality of agents.

2. The method of claim 1 , wherein tagging the message with the set of features further comprises:
tagging the message with the set of features in real-time as the dialog is progressing.

3. The method of claim 1 , wherein tagging the message with the set of features further comprises:
tagging the message with a feature that indicates a sequence aspect of the message in the dialog.

4. The method of claim 1 , wherein tagging the message with the set of features further comprises tagging the message with the set of features by:
checking for a presence of a set of canned phrases in a specified message in the storage system to determine a score of the message, the set of canned phrases comprising a group of phrases having a frequency of usage by the agent that is above a first specified threshold;
identifying a canned phrase of the set of canned phrases for which the score of the message is above a second specified threshold; and
tagging the message with a feature of the canned phrase.

5. The method of claim 4 , wherein the feature of the set of canned phrases indicates a stage in a progress of the dialog.

6. The method of claim 4 further comprising:
receiving, by the computing system, a portion of an input message that is input on a first computing device associated with the agent;
generating, by the computing system, a suggested message on the first computing device for use by the agent to complete a remaining portion of the input message, wherein the suggested message generated is based on a comparison of the portion of the input message with the canned phrase or predefined set of vocabulary of words or phrases.

7. The method of claim 6 further comprising:
tagging, by the computing system, the suggested message with a feature that indicates whether the agent has accepted the suggested message.

8. The method of claim 1 , wherein tagging the message with the set of features further comprises:
analyzing the message to determine an emotion associated with the message and tagging the message with a feature depicting the emotion.

9. The method of claim 8 , wherein tagging the message to determine the emotion further comprises:
identifying the emotion based on at least one of a plurality of words, a plurality of symbols, or a plurality of emoticons in the message, and/or moods, context from other messages and like/dislike tags from the customer, the agent, supervisor, other agents, or a QA team.

10. The method of claim 1 , wherein tagging the message with the set of features further comprises:
analyzing the message to determine a commitment made by the agent to the customer and tagging the message with a feature depicting the commitment.

11. The method of claim 10 , wherein analyzing the message to determine the commitment further comprises:
identifying the commitment based on a plurality of words in the message that indicate the commitment.

12. The method of claim 1 , wherein tagging the message with the set of features further comprises:
analyzing the message to determine a stage of the dialog and tagging the message with a feature depicting the stage.

13. The method of claim 12 , wherein analyzing the message to determine the stage of the dialog further comprises:
inferring the stage based on a plurality of words used in the message.

14. The method of claim 1 , wherein tagging the message with the set of features further comprises:
analyzing the message to determine an intent of the dialog and tagging the message with a feature depicting the intent.

15. The method of claim 14 , wherein analyzing the message to determine the intent of the dialog further comprises:
inferring or classifying the intent based on a plurality of words used in the message.

16. The method of claim 1 , wherein tagging the message with the set of features further comprises:
tagging, by the computing system, the message with a feature input by at least one of the customer, the agent or a supervisor of the agent.

17. The method of claim 1 , wherein the set of tagged features comprises at least one of an intent feature, an emotion feature, a commitment feature, or a stage confirmation feature.

18. The method of claim 17 , wherein the intent feature is input by a user selection of an entry from a list of entries generated by the computing system;
wherein each of entries indicates an intent of the customer for contacting the agent;
wherein the list of entries is generated at least as a function of any of:
the plurality of messages of the dialog;
the set of features of the plurality of messages;
the second plurality of messages;
the one or more features associated with the second plurality of messages; and
a transaction record associated with the customer.

19. The method of claim 16 , wherein the feature input by at least one of the agent or the supervisor comprises a hash tag feature that indicates a context of the message in the dialog.

20. The method of claim 16 , wherein the feature input by at least one of the agent or the supervisor further comprises:
a collaborative label feature that indicates an aspect of a recommendation message generated by the computing system, wherein the recommendation message is used by the agent as a next response to the customer that progresses the dialog towards a specified goal.

21. The method of claim 16 , wherein the feature input by at least one of the agent or the supervisor further comprises:
a best response feature indicating that a recommendation message generated by the computing system helped the agent in advancing the dialog towards a favorable resolution, wherein the recommendation message is used by the agent as a response to the customer at any appropriate stage of the rest of the dialog to advance the dialog towards the favorable resolution.

22. The method of claim 21 further comprising:
sharing one or more of the plurality of messages that are tagged with the best response feature with the plurality of agents.

23. The method of claim 1 further comprising:
executing, by the computing system, a text mining process on the plurality of messages, the second plurality of messages, and a transaction record associated with the customer to extract a specified set of features and a specified set of messages as mined data;
generating, by the co

Assignees

• [24]7 Ai Inc

Inventors

• Seth Suchana
• Singh Bhupinder
• Ramachandran Mathangi Sri
• Vijayaraghavan Ravi
• Kannan Pallipuram V

Classifications

• G06Q30/02

  Marketing; Price estimation or determination; Fundraising · CPC title

• G06F16/3329

  Natural language query formulation · CPC title

• G06Q30/016

  After-sales · CPC title

• G06F16/3322 Primary

  using system suggestions (G06F16/3325 takes precedence) · CPC title

• H04L67/10

  in which an application is distributed across nodes in the network (software deployment G06F8/60; multiprogramming arrangements G06F9/46) · CPC title