Method, computer-readable storage device, and apparatus for analyzing text messages

• US 10,453,079 B2
• Filed: 11/20/2013
• Issued: 10/22/2019
• Est. Priority Date: 11/20/2013
• Status: Expired due to Fees

• Current Assignee: AT&T Intellectual Property I LP (AT&T, Inc.)
• Original Assignee: AT&T Intellectual Property I LP (AT&T, Inc.)
• Inventors: Bodden, Christopher David
• Primary Examiner(s): Waesco, Joseph M
• Application Number:
  US14/085,667
• Publication Number:
  US 20150142510A1
• Time in Patent Office: 2,162 Days
• Field of Search: None
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data
• CPC Class Codes: H04W 4/14 Short messaging services, e.g. short message services [SMS] or unstructured supplementary service data [USSD]

Abstract

A method, computer-readable storage device and apparatus for analyzing text messages are disclosed. For example, the method receives the text messages, tokenizes each one of the text messages to extract a token, identifies the token as trending, wherein trending comprises the token having a frequency above a threshold, classifies a sentiment for each one of the text messages having the token that is identified as trending, identifies a demographic for the sentiment of the token that is trending, and provides the demographic and the sentiment of the token that is trending to a third party entity associated with the token.

Claims

1. A method for analyzing text messages, the method comprising:

receiving, by a processor of a telecommunication service provider, over a communication network of the telecommunication service provider, the text messages;

storing, by the processor, to a memory, the text messages;

tokenizing, by the processor, each one of the text messages to extract a token, wherein the token relates to a product or a service of a third party entity;

identifying, by the processor, the token as trending, wherein the trending comprises the token having a frequency of occurrence in the text messages within a period of time that is above a threshold;

classifying, by the processor in response to the identifying the token as trending, a sentiment for each one of the text messages having the token that is identified as trending, wherein the sentiment is classified using a naive bayes classifier;

identifying, by the processor, a demographic for each one of the text messages having the token that is identified as trending, wherein the demographic comprises a relationship type between a sender and a recipient of each one of the text messages having the sentiment of the token that is identified as trending, wherein the demographic is determined from subscriber information of the telecommunication service provider, wherein the relationship type is identified based upon a group subscription plan subscribed to by the sender and the recipient, wherein the relationship type is at least one of;

a parent to children relationship, a spouse to spouse relationship, an employer to employee relationship, a supervisor to employee relationship, or an employee to a particular department or business unit relationship;

removing, by the processor, information related to an identity of the sender and an identity of the recipient from each one of the text messages having the token that is identified as trending in response to the identifying the demographic;

generating, by the processor, an overall sentiment for the token that is identified as trending; and

providing, by the processor, the overall sentiment, or the demographic and the sentiment for each one of the text messages having the token that is identified as trending to an endpoint device of the third party entity associated with the token, wherein the demographic and the sentiment of the token that is identified as trending are presented in a list based on the relationship type for allowing the third party entity to distinguish the sentiment of the product or the service with respect to family members and company employees.

2. The method of claim 1, wherein the tokenizing comprises:

applying, by the processor, a natural language processor to identify a potential token in each of the text messages; and

comparing, by the processor, the potential token to a pre-defined library to confirm that the potential token is the token.

3. The method of claim 1, wherein the sentiment comprises a positive sentiment.

4. The method of claim 1, wherein the sentiment comprises a negative sentiment.

5. The method of claim 1, wherein the generating the overall sentiment comprises a comparison of a percentage of positive sentiments versus a percentage of negative sentiments of the text messages having the token that is identified as trending.

6. The method of claim 1, wherein the group subscription plan comprises a family plan.

7. The method of claim 1, wherein the group subscription plan comprises a corporate plan.

8. The method of claim 1, wherein the third party entity pays for a subscription to be provided with the demographic and the sentiment of the token that is identified as trending.

9. A non-transitory computer-readable storage device storing a plurality of instructions which, when executed by a processor of a telecommunication service provider, cause the processor to perform operations for analyzing text messages, the operations comprising:

receiving, over a communication network of the telecommunication service provider, the text messages;

storing, by the processor, to a memory, the text messages;

tokenizing each one of the text messages to extract a token, wherein the token relates to a product or a service of a third party entity;

identifying the token as trending, wherein the trending comprises the token having a frequency of occurrence in the text messages within a period of time that is above a threshold;

classifying, in response to the identifying the token as trending, a sentiment for each one of the text messages having the token that is identified as trending, wherein the sentiment is classified using a naive bayes classifier;

identifying a demographic for each one of the text messages having the token that is identified as trending, wherein the demographic comprises a relationship type between a sender and a recipient of each one of the text messages having the sentiment of the token that is identified as trending, wherein the demographic is determined from subscriber information of the telecommunication service provider, wherein the relationship type is identified based upon a group subscription plan subscribed to by the sender and the recipient, wherein the relationship type is at least one of;

a parent to children relationship, a spouse to spouse relationship, an employer to employee relationship, a supervisor to employee relationship, or an employee to a particular department or business unit relationship;

removing information related to an identity of the sender and an identity of the recipient from each one of the text messages having the token that is identified as trending in response to the identifying the demographic;

generating an overall sentiment for the token that is identified as trending; and

providing the overall sentiment, or the demographic and the sentiment for each one of the text messages having the token that is identified as trending to an endpoint device of the third party entity associated with the token, wherein the demographic and the sentiment of the token that is identified as trending are presented in a list based on the relationship type for allowing the third party entity to distinguish the sentiment of the product or the service with respect to family members and company employees.

10. The non-transitory computer-readable storage device of claim 9, wherein the tokenizing comprises:

applying a natural language processor to identify a potential token in each of the text messages; and

comparing the potential token to a pre-defined library to confirm that the potential token is the token.

11. The non-transitory computer-readable storage device of claim 9, wherein the generating the overall sentiment comprises a comparison of a percentage of positive sentiments versus a percentage of negative sentiments of the text messages having the token that is identified as trending.

12. The non-transitory computer-readable storage device of claim 9, wherein the third party entity pays for a subscription to be provided with the demographic and the sentiment of the token that is trending.

13. An apparatus for analyzing text messages, the apparatus comprising:

a processor of a telecommunication service provider; and

a computer-readable storage device storing a plurality of instructions which, when executed by the processor, cause the processor to perform operations, the operations comprising;

receiving, over a communication network of the telecommunication service provider, the text messages;

storing, to a memory, the text messages;

tokenizing each one of the text messages to extract a token, wherein the token relates to a product or a service of a third party entity;

identifying the token as trending, wherein the trending comprises the token having a frequency of occurrence in the text messages within a period of time that is above a threshold;

classifying, in response to the identifying the token as trending, a sentiment for each one of the text messages having the token that is identified as trending, wherein the sentiment is classified using a naive bayes classifier;

identifying a demographic for each one of the text messages having the token that is identified as trending, wherein the demographic comprises a relationship type between a sender and a recipient of each one of the text messages having the sentiment of the token that is identified as trending, wherein the demographic is determined from subscriber information of the telecommunication service provider, wherein the relationship type is identified based upon a group subscription plan subscribed to by the sender and the recipient, wherein the relationship type is at least one of;

a parent to children relationship, a spouse to spouse relationship, an employer to employee relationship, a supervisor to employee relationship, or an employee to a particular department or business unit relationship;

removing information related to an identity of the sender and an identity of the recipient from each one of the text messages having of the token that is identified as trending in response to the identifying the demographic;

generating an overall sentiment for the token that is identified as trending; and

providing the overall sentiment, or the demographic and the sentiment for each one of the text messages having the token that is identified as trending to an endpoint device of the third party entity associated with the token, wherein the demographic and the sentiment of the token that is identified as trending are presented in a list based on the relationship type for allowing the third party entity to distinguish the sentiment of the product or the service with respect to family members and company employees.

14. The apparatus of claim 13, wherein the tokenizing comprises:

applying a natural language processor to identify a potential token in each of the text messages; and

comparing the potential token to a pre-defined library to confirm that the potential token is the token.

15. The apparatus of claim 13, wherein the sentiment comprises a positive sentiment.

16. The apparatus of claim 13, wherein the sentiment comprises a negative sentiment.

17. The apparatus of claim 13, wherein the generating the overall sentiment comprises a comparison of a percentage of positive sentiments versus a percentage of negative sentiments of the text messages having the token that is identified as trending.

18. The apparatus of claim 13, wherein the group subscription plan comprises a family plan.

19. The apparatus of claim 13, wherein the group subscription plan comprises a corporate plan.

20. The apparatus of claim 13, wherein the third party entity pays for a subscription to be provided with the demographic and the sentiment of the token that is identified as trending.