Methods, systems, and articles of manufacture for analyzing social media with trained intelligent systems to enhance direct marketing opportunities

• US 9,710,829 B1
• Filed: 06/19/2013
• Issued: 07/18/2017
• Est. Priority Date: 06/19/2013
• Status: Active Grant

• Current Assignee: Intuit, Inc.
• Original Assignee: Intuit, Inc.
• Inventors: Chen, Anqi, Reeves, Daniel Edward, Kulkarni, Ravindra Bindumadhav, Ojha, Preetam Kumar, Sitapara, Ojas R.
• Primary Examiner(s): Sittner, Matthew T
• Application Number:
  US13/922,107
• Time in Patent Office: 1,490 Days
• CPC Class Codes: G06Q 30/0276 Advertisement creation
• CPC Class Codes: G06Q 50/01 Social networking

Abstract

Disclosed are methods, systems, and articles of manufactures for analyzing user generated content items in social media networks with trained intelligent systems to create or enhance direct marketing opportunities. The method or the system monitors user generated content items in social media networks and identifies a relevant user generated content item that may be materialized into a direct marketing opportunity. The method or system further performs language processing on the relevant user generated content item and uses the processing results to prepare a response which is subsequently transmitted to the user to materialize the direct marketing opportunity. The method or system uses various intelligent logic processes or modules that may be further enhanced by machine learning techniques with human expert reviews and intervention to improve their respective accuracy, reliability, or confidence level.

Claims

1. A computer implemented method for analyzing user generated content items in social media networks with trained intelligent systems, comprising:

at least one computing system identifying a user generated content item that is transmitted from a user computing device of a user in a social media network via a first network element, wherein the user generated content item includes at least a part that is expressed in a natural language;

the at least one computing system determining whether the user generated content item is to be further processed with additional processing at least by performing a first filtering process based in part or in whole upon a set of key terms or a set of filtering rules, wherein user generated content items determined not to be relevant are discarded from the additional processing;

when the user generated content item is determined to be relevant, the additional processing comprising;

performing, at an artificial intelligence module stored in memory and including or functioning in conjunction with at least one micro-processor of the at least one computing system, a segmentation process on the user generated content item by segmenting at least a part of the user generated content item into a plurality of units, storage of the plurality of units in a first location of memory of the computing system, and a second filtering process on the user generated content item, the second filtering process producing a smaller user generated content item by discarding a smaller objective portion of the user generated content item;

identifying or creating, with at least the artificial intelligence module, a response for the user generated content item at least by referencing at least results of one or more analyses of the plurality of units stored in a second location of the memory and corresponding to the smaller user generated content item in light of an environment in which the plurality of units are used and by calibrating the artificial intelligence module via at least validating the user generated content item, wherein the response includes at least a portion that is expressed in the natural language, and the one or more analyses and validating the user generated content item enhance accuracy of description of the user generated content item determined by the artificial intelligence module;

the at least one computing system storing the response in a database and indexing the response with a database index that facilitates subsequent retrievals of the response to subsequently identified user generated content items;

the at least one computing system transmitting, via a second network element, the response from the at least one computing system to a user computing device of the user in response to the user generated content item; and

evaluating the response for the user generated content item at least by receiving user input data responding to the response from the user computing device and by reducing the user input data into reduced user input data, wherein reducing the user input data comprises filtering out one or more true elements, and analyzing the reduced user input data based in part or in whole upon a context in which the reduced user input data appears.

2. The computer implemented method of claim 1, further comprising the at least one computing system performing a post-response sentiment analysis.

3. The computer implemented method of claim 1, in which the at least one computing system identifying or creating the response for the user generated content item further comprises:

the at least one computing system identifying or receiving a plurality of user generated content items from the social media network; and

the at least one computing system determining the relevant user generated content item by performing a process on the plurality of user generated content items.

4. The computer implemented method of claim 3, in which the process performed on the plurality of user generated content items comprises a filtering process to identify a subjective or relevant portion from at least some of the plurality of user generated content items or to remove an objective or irrelevant portion from the at least some of the plurality of user generated content items.

5. The computer implemented method of claim 3, in which the filtering process removes one or more objective elements from the at least some of the plurality of user generated content items, and one or more objective elements comprise one or more universally true statements.

6. The computer implemented method of claim 3, in which the at least one computing system that identifying or creating the response for the user generated content item further comprises:

the at least one computing system forwarding data or information associated with the response or the relevant user generated content item for validation by a human expert; and

the at least one computing system adjusting the process based at least in part upon a result of the validation by the human expert.

7. The computer implemented method of claim 1, in which the at least one computing system identifying or creating the response for the user generated content item further comprises at least one of:

the at least one computing system classifying the user generated content item by using at least one of a syntactic natural language processor, support vector machines, and a Bayesian classifier including a Naï

ve Bayes classification process;

the at least one computing system indexing the user generated content item in one or more data structures; and

the at least one computing system indexing the response to the user generated content item in the one or more data structures or in other one or more data structures.

8. The computer implemented method of claim 1, in which the at least one computing system identifying the at least one computing system identifies or create the response for the user generated content item comprises at least one computing system determining one or more relations between the response and one or more products, one or more services, or a combination of the one or more products and the one or more services.

9. The computer implemented method of claim 1, in which the at least one computing system identifying the at least one computing system identifies or create a response for the user generated content item comprises at least one of:

at least one computing system performing a process on one or more existing responses;

at least one computing system indexing the on one or more existing responses;

at least one computing system validating at least some of the one or more existing responses;

at least one computing system classifying the one or more existing responses;

at least one computing system identifying one or more templates associated with the social media network;

at least one computing system identifying one or more presentation requirements associated with the social media network; and

at least one computing system performing one or more statistical analyses using data or information associated with the response.

10. The computer implemented method of claim 1, in which the at least one computing system identifying or creating the response for the user generated content item further comprises the at least one computing system understanding an item by performing a natural language processing on the item.

11. The computer implemented method of claim 10, in which the natural language processing performed on comprises at least one of a lexical analysis, a syntax analysis, a semantics analysis, a context analysis, a data mining process, a classification process, a data mining process, a data dredging process, an assisted machine learning process, and an unassisted machine learning process.

12. The computer implemented method of claim 10, in which the item comprises the user generated content item, the response, or a user input.

13. The computer implemented method of claim 1, in which the at least one computing system transmitting the response to the user in response to the user generated content item comprises at least one of:

the at least one computing system providing an interactive list of topics related to one or more products or one or more services in the response;

the at least one computing system performing a fuzzy matching between a user input and data or information related to the one or more products or one or more services;

the at least one computing system presenting a list of questions to the user to inquire into effectiveness or helpfulness of the response or user'"'"'s knowledge;

the at least one computing system providing a step-by-step guide, a trial version of the one or more products or one or more services, or an option for the user to request for live support; and

the at least one computing system monitoring progress of the response after the review is transmitted to be reviewed by the user.

14. The computer implemented method of claim 1, further comprising the at least one computing system performing a post-response evaluation or improvement process after the response has been transmitted for the user to review.

15. The computer implemented method of claim 1, in which the at least one computing system performing the post-response evaluation or improvement process further comprising:

the at least one computing system identifying a deviation between information associated with the user generated content item and available information associated with a product, a service, or a combination of the product and the service;

the at least one computing system identifying a cause for the deviation between the user generated content item and the product, the service, or the combination of the product and the service; and

the at least one computing system identifying an area for improvement for the product, the service, or the combination of the product and the service.

16. The computer implemented method of claim 1, in which the at least one computing system performing the post-response evaluation or improvement process further comprising:

the at least one computing system performing a user sentiment analysis process;

the at least one computing system improving accuracy of identifying or creating the response in response to the user generated content item based at least in part upon a result of the user sentiment analysis; and

creating or generating another response with improved accuracy.

17. The computer implemented method of claim 1, in which the at least one computing system performing the user sentiment analysis process comprising at least one of:

the at least one computing system evaluating data related to the response or user'"'"'s feedback;

the at least one computing system performing a language processing on the data;

the at least one computing system removing a portion of the data from the data; and

the at least one computing system extracting another portion of the data from the data.

18. The computer implemented method of claim 17, in which the at least one computing system performing the user sentiment analysis process comprising at least one of:

the at least one computing system determining or classifying polarity of the user sentiment;

the at least one computing system improving accuracy of the user sentiment analysis by using an assisted or unassisted machine learning technique; and

the at least one computing system determining a feature or an aspect of a product or a service on which the user sentiment is expressed; and

the at least one computing system determining whether a follow-up response is to be sent for the user to review based at least in part upon the user sentiment.

19. The computer implemented method of claim 1, in which the at least one computing system that is used to generate the response also hosts the product or the service.

20. The computer implemented method of claim 1, further comprising an intermediary computing system hosting the product or the service, wherein the intermediary computing system is operatively connected to the at least one computing system that is used to generate the response.

21. An article of manufacture comprising a non-transitory computer readable storage medium storing thereupon a sequence of instructions which, when executed by at least one computing system having at least one processor, causes the at least one processor to perform a set of acts for analyzing user generated content items in social media networks with trained intelligent systems, comprising:

a first computer program code, a compiled version of which, when executed by the at least one computing system, causes the at least one computing system to;

identify a user generated content item that is transmitted from a user computing device of a user in a social media network via a first network element, wherein the user generated content item includes at least a part that is expressed in a natural language;

the at least one computing system determining whether the user generated content item is to be further processed with additional processing at least by performing a first filtering process based in part or in whole upon a set of key terms or a set of filtering rules, wherein user generated content items determined not to be relevant are discarded from the additional processing;

when the user generated content item is determined to be relevant, the first computer program code further cause the at least one computer system to perform the additional processing to;

perform, at an artificial intelligence module stored in memory and including or functioning in conjunction with at least one micro-processor of the at least one computing system, a segmentation process on the user generated content item by segmenting at least a part of the user generated content item into a plurality of units, storage of the plurality of units in a first location of memory of the computing system, and a second filtering process on the user generated content item, the second filtering process producing a smaller user generated content item by discarding a smaller objective portion of the user generated content item;

identify or create, with at least the artificial intelligence module, a response for the user generated content item at least by referencing at least results of one or more analyses of the plurality of units stored in a second location of the memory and corresponding to the smaller user generated content item in light of an environment in which the plurality of units are used and by calibrating the artificial intelligence module via at least validating the user generated content item, wherein the response includes at least a portion that is expressed in the natural language, and the one or more analyses and validating the user generated content item enhance accuracy of description of the user generated content item determined by the artificial intelligence module;

store the response in a database and index the response with a database index that facilitates subsequent retrievals of the response to subsequently identified user generated content items;

transmit, via a second network element, the response from the at least one computing system to a user computing device of the user in response to the user generated content item; and

evaluate the response for the user generated content item at least by receiving user input data responding to the response from the user computing device and by reducing the user input data into reduced user input data, wherein reducing the user input data comprises filtering out one or more true elements, and analyzing the reduced user input data based in part or in whole upon a context in which the reduced user input data appears.

22. The article of manufacture of claim 21, further comprising:

a second computer program code, a compiled version of which, when executed by the at least one computing system, causes the at least one computing system to perform a post-response sentiment analysis.

23. The article of manufacture of claim 21, in which the first program code that is to cause the at least one computing system to identify or create the response for the user generated content item further comprises:

a second computer program code, a compiled version of which, when executed by the at least one computing system, causes the at least one computing system to;

identify or receive a plurality of user generated content items from the social media network; and

determine the relevant user generated content item by performing a process on the plurality of user generated content items.

24. The article of manufacture of claim 23, in which the compiled version of the second computer program code, when executed by the at least one computing system, causes the at least one computing system to:

forward data or information associated with the response or the relevant user generated content item for validation by a human expert; and

adjust the process based at least in part upon a result of the validation by the human expert.

25. The article of manufacture of claim 21, in which the first program code that is to cause the at least one computing system to identify or create the response for the user generated content item further comprises at least one of:

a second computer program code, a compiled version of which, when executed by the at least one computing system, causes the at least one computing system to;

classify the user generated content item by using at least one of a syntactic natural language processor, support vector machines, and a Bayesian classifier including a Naï

ve Byaes classification engine;

index the user generated content item in one or more data structures; and

index the response to the user generated content item in the one or more data structures or in other one or more data structures.

26. The article of manufacture of claim 21, in which the first program code that is to cause the at least one computing system to identify or create the response for the user generated content item comprises:

a second computer program code, a compiled version of which, when executed by the at least one computing system, causes the at least one computing system to;

determine one or more relations between the response and one or more products, one or more services, or a combination of the one or more processes and the one or more services.

27. The article of manufacture of claim 21, in which the first program code that is to cause the at least one computing system to identify or create a response for the user generated content item comprises at least one of:

a second computer program code, a compiled version of which, when executed by the at least one computing system, causes the at least one computing system to;

perform a process on one or more existing responses;

index the on one or more existing responses;

validate at least some of the one or more existing responses;

classify the one or more existing responses;

identify one or more templates associated with the social media network;

identify one or more presentation requirements associated with the social media network; and

perform one or more statistical analyses using data or information associated with the response.

28. A system for analyzing user generated content items in social media networks with trained intelligent systems, comprising:

memory storing thereupon a program code;

a first computing system that comprises at least one processor and is configured to execute a sequence of instructions for executing the program code in conjunction with at least;

an artificial intelligent module executes the sequence of instructions to perform to identify a user generated content item that is transmitted from a user computing device of a user in a social media network via a first network element, wherein the user generated content item includes at least a part that is expressed in a natural language;

the first computing system executes the sequence of instructions to determine whether the user generated content item is to be further processed with additional processing at least by performing a first filtering process based in part or in whole upon a set of key terms or a set of filtering rules, wherein user generated content items determined not to be relevant are discarded from the additional processing;

when the user generated content item is determined to be relevant, the first computing system further performs the additional processing to invoke;

the artificial intelligent module stored in memory and including or functioning in conjunction with at least one micro-processor of the first computing system to execute the sequence of instructions to perform a segmentation process on the user generated content item by segmenting at least a part of the user generated content item into a plurality of units, storage of the plurality of units in a first location of the memory of the first computing system, and a second filtering process on the user generated content item, the second filtering process producing a smaller user generated content item by discarding a smaller objective portion of the user generated content item;

the artificial intelligence module to further execute the sequence of instructions to identify or create a response for the user generated content item at least by referencing at least results of one or more analyses of the plurality of units stored in a second location of the memory and corresponding to the smaller user generated content item in light of an environment in which the plurality of units are used and by calibrating the artificial intelligence module via at least validating the user generated content item, wherein the response includes at least a portion that is expressed in the natural language, and the one or more analyses and validating the user generated content item enhance accuracy of description of the user generated content item determined by the language processing mechanism;

an execution of the sequence of instructions to store the response in a database and index the response with a database index that facilitates subsequent retrievals of the response to subsequently identified user generated content items;

an execution of the sequence of instructions to transmit, via a second network element, the response from the first computing system to a user computing device of the user in response to the user generated content item; and

an execution the sequence of instructions to evaluate the response for the user generated content item at least by receiving user input data responding to the response from the user computing device and by reducing the user input data into reduced user input data, wherein reducing the user input data comprises filtering out one or more true elements, and analyzing the reduced user input data based in part or in whole upon a context in which the reduced user input data appears.

29. The system of claim 28, in which the first computing system is further to work in conjunction with a sentiment analysis module to perform a post-response sentiment analysis.

30. The system of claim 28, in which the first computing system that executes the sequence of instructions to identify or create the response for the user generated content item is further to:

identify or receive a plurality of user generated content items from the social media network; and

determine the relevant user generated content item by performing a process on the plurality of user generated content items.

31. The system of claim 30, in which the first computing system further executes the sequence of instructions to:

forward data or information associated with the response or the relevant user generated content item for validation by a human expert; and

adjust the process based at least in part upon a result of the validation by the human expert.

32. The system of claim 28, in which the first computing system further executes the sequence of instructions to:

classify the user generated content item by using one or more of support vector machines, a syntactic natural language processing module, or a Bayesian classifier which further includes a Naï

ve Bayes module;

index the user generated content item in one or more data structures;

or

index the response to the user generated content item in the one or more data structures or in other one or more data structures.

33. The system of claim 28, in which the at least one computing system that executes the sequence of instructions to identify or create the response for the user generated content item is further to determine one or more relations between the response and one or more products, one or more services, or a combination of the one or more processes and the one or more services.

34. The system of claim 28, in which the at least one computing system that executes the sequence of instructions to identify or create a response for the user generated content item is further to:

perform a process on one or more existing responses;

index the on one or more existing responses;

validate at least some of the one or more existing responses;

classify the one or more existing responses;

identify one or more templates associated with the social media network;

identify one or more presentation requirements associated with the social media network;

or

perform one or more statistical analyses using data or information associated with the response.