System and method for auto-suggesting responses based on social conversational contents in customer care services

• US 10,162,884 B2
• Filed: 07/23/2013
• Issued: 12/25/2018
• Est. Priority Date: 07/23/2013
• Status: Active Grant

• Current Assignee: Conduent Business Services, LLC (Conduent, Inc.)
• Original Assignee: Conduent Business Services, LLC (Conduent, Inc.)
• Inventors: Thirugnanasundaram, Veerasundaravel, Vandervort, David R., Sun, Tong, Bakthavachalu, Arun
• Primary Examiner(s): Perveen, Rehana
• Application Number:
  US13/948,204
• Publication Number:
  US 20150032724A1
• Time in Patent Office: 1,981 Days
• Field of Search: 707722
• CPC Class Codes: G06F 16/3329 Natural language query formulation or dialogue systems
• CPC Class Codes: G06Q 10/10 Office automation; Time management
• CPC Class Codes: G06Q 30/01 Customer relationship services
• CPC Class Codes: H04L 51/02 using automatic reactions or user delegation, e.g. automatic replies or chatbot-generated messages
• CPC Class Codes: H04L 51/216 Handling conversation history, e.g. grouping of messages in sessions or threads
• CPC Class Codes: H04L 51/52 for supporting social networking services

Abstract

A first embodiment of the disclosure relates to a method for responding to a message posted in a social media stream. The method includes monitoring a social media site for at least one message including select subject matter. In response to identifying a message, the method includes collecting a series of exchanges that form a conversational thread including the message. The method includes determining at least one content attribute of the message. The method includes classifying the message using at least one key attribute. The method includes searching a database for a reference message using a combination of the at least one content and key attributes. The method includes determining a previous outcome of a reference thread including the reference message. The method includes using the previous outcome for determining a course of action.

Claims

1. A computer-implemented method for responding to a current message posted in a social media stream, the method comprising:

at a server computer, implementing instructions for;

monitoring a social media site for at least one current message including select subject matter;

in response to identifying a current message, collecting a series of exchanges that form a conversational thread including the current message;

determining at least one content attribute of the conversational thread including the current message, including;

extracting a set of tags forming the current message by;

splitting at least the current message into separate words;

filtering and removing irrelevant words so that only contextual information remains;

associating each of the remaining words as a tag,

forming at least one combination of tags to generate at least one slug, and

associating each slug as a content attribute;

classifying the current message using at least one key attribute;

searching a database for a reference message using a combination of the at least one content and key attributes;

determining a previous outcome of a reference thread including the reference message, the determining including;

parsing words in a customer-posted response of the reference thread, where the customer-posted response was made after a user'"'"'s previous solution was posted to the reference thread, and

automatically applying the parsed words to a stored dictionary to associate the words as belonging to one of a number of predetermined sentiments; and

,

in response to the words belonging to a positive sentiment, displaying the user'"'"'s previous solution for determining a course of action.

2. The method of claim 1, wherein the at least one content attribute is selected from a group consisting of:

a set of tags extracted from the current message;

at least one slug generated from at least one combination of tags; and

,

quoted text of the current message.

3. The method of claim 1, wherein the key attribute is selected from a group consisting of:

a category;

a topic that the current message relates to;

a location; and

,

a combination of the above.

4. The method of claim 1 further comprising:

rating a confidence of the reference message based on the combination of the at least one content and key attributes used in the searching.

5. The method of claim 1, further comprising:

categorizing the current message into one of a predetermined set of categories classifying a purpose of the message; and

,

searching the database using at least the category one key attribute in the combination.

6. The method of claim 1 further comprising:

comparing tags in the current message with a stored dictionary for identifying at least one topic of the current message;

classifying the current message as belonging to the topic; and

,

searching the database for the reference message using at least the topic as one key attribute in the combination.

7. The method of claim 1 further comprising:

determining a location of a device posting the current message; and

searching the database for the reference message using at least the location as one key attribute in the combination.

8. The method of claim 1, wherein the determining the course of action includes:

accessing a database of reference threads to determine if the current message corresponds to one of a set of reference threads;

for the current message being determined as corresponding to a reference thread, identifying a sentiment of a last message in the reference thread as belonging to one of a negative and a positive sentiment;

for the message being determined as not corresponding to a reference thread, providing the message to a user for review.

9. The method of claim 5, wherein the determining the course of action includes:

for the reference message belonging to the positive sentiment, providing the previous outcome to the user as a suggested response to the thread; and

,

for the reference message belonging to the negative sentiment, providing the message to the user for review.

10. The method of claim 5, wherein the reference threads include previous threads that were concluded.

11. A computer program product comprising non-transitory tangible media which encodes instructions for performing the method of claim 1.

20. The method of claim 1, wherein the extracting of the tags includes:

splitting the conversational thread into separate words;

parsing the split words to generate stem words.

12. A system for responding to a message posted in a social media stream, the system comprising:

an auto-suggesting response device comprising;

a memory;

a downloadable software module that operates on a server computer linked to a customer computing device by a network;

a processor coupled to the memory programmed with executable instructions for;

monitoring a social media site for at least one current message including select subject matter, and,

in response to identifying a current message, collecting a series of exchanges that form a conversational thread including the message,

extracting a set of tags forming the current message, the extracting including;

splitting at least the current message into separate stem words,

filtering and removing irrelevant words from the separated words so that only contextual information remains, and

associating each of the remaining words as a tag,

forming at least one combination of tags to generate at least one slug, and associating each slug as a content attribute;

categorizing the current message into one of a predetermined set of categories,

searching a database for a reference message using a combination of the at least one content and key attributes, where in the content attribute is extracted from the conversational thread,

determining an outcome of a reference thread including the reference message by parsing words in a previous customer-posted response of the reference thread, where the customer-posted response was made received after a user'"'"'s previous solution was posted to the reference thread and automatically applying the parsed words to a stored dictionary to associate the previous customer-posted response as belonging to one of a number of predetermined sentiments, and

in response to the previous customer-posted response belonging to a positive sentiment, displaying the user-s previous solution for providing in the thread.

13. The system of claim 12, wherein the predetermined set of categories includes at least one of:

reasons why messages can be posted in the thread;

types of sentiments that the message conveys based on tags in the message;

topics of messages;

locations of messages; and

,

a combination of the above.

14. The system of claim 12, wherein the processor executes further instructions for:

comparing tags in the current message with a stored dictionary for identifying at least one topic of the message;

assigning the topic to the current message; and

,

searching the database for the reference message using the topic.

15. The system of claim 12, wherein the processor executes further instructions for:

determining a location of a device posting the current message, and

searching the database for the reference message using the location.

16. The system of claim 12, wherein the processor executes further instructions for:

categorizing the current message into one of the predetermined set of categories based on the slugs; and

,

searching the database for the reference message by comparing the slugs with reference messages.

17. The system of claim 12, wherein the processor executes further instructions for:

in response to the response search module accessing a database of reference threads to determine if the current message corresponds to one of a set of reference threads, for the current message being determined as corresponding to a reference thread, identifying a sentiment of a last message in the reference thread as belonging to one of a negative and a positive sentiment.

18. The system of claim 15, wherein the processor executes further instructions for:

for the reference message belonging to the positive sentiment, providing an answer posted to the reference thread as a reply to the current message; and

,

for the reference message belonging to the negative sentiment, providing the message to the user for review.

19. The system of claim 16, wherein the reference threads include previous threads that were concluded.