Method and system for generating social signal vocabularies

• US 9,959,548 B2
• Filed: 07/21/2014
• Issued: 05/01/2018
• Est. Priority Date: 08/31/2012
• Status: Active Grant

• Current Assignee: TPG Specialty Lending Incorporated
• Original Assignee: Sprinklr Inc.
• Inventors: Huddleston, Erik Lee, Huddleston, Brian, Potter, Tim
• Primary Examiner(s): Lazaro, David
• Assistant Examiner(s): SHITAYEWOLDETSADI, BERHANU
• Application Number:
  US14/336,914
• Publication Number:
  US 20140330632A1
• Time in Patent Office: 1,380 Days
• Field of Search: 709206-245
• CPC Class Codes: G06Q 30/0242 Determining effectiveness of advertisements
• CPC Class Codes: G06Q 50/01 Social networking
• CPC Class Codes: H04B 7/26 at least one of which is mobile
• 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
• CPC Class Codes: H04L 65/403 Arrangements for multi-party communication, e.g. for conferences data switching systems for conference H04L12/18; arrangements for connecting several subscribers to a common circuit, i.e. affording conference facilities H04M3/56; television conferencing systems H04N7/15

Abstract

A social analytic system may identify the social signals associated with a brand, campaign, or any other topic. The social analytic system may generate a vocabulary associated with the brand, campaign, or topic based terms used in the associated social signals. The vocabulary may be used for generating social media analytics and identifying social media events, such as marketing campaigns. In one example, a vocabulary may be compared with vocabularies associated with different constituents to identify the positive and negative terms in the vocabulary.

Claims

1. A method, comprising:

identifying, by a computing device, a set of social signals generated on an electronic social network;

identifying, by the computing device and at a first time, a first subset of said set of social signals generated on the electronic social network;

wherein the social signals of the first subset are associated with a campaign defined by electronically stored campaign data, wherein the campaign data is associated with known campaign terminology;

inputting the social signals of the first subset into a natural language processor to discover new campaign terminology that is different than the known campaign terminology, the new campaign terminology not a priori associated with the campaign by the campaign data, wherein the new campaign terminology is discovered at a second time that is later than the first time;

identifying a first keyword based on the new campaign terminology;

identifying a second additional subset of said set of social signals using the first keyword, wherein the second additional subset is different than the first subset, wherein the identifying the second additional subset is by the computing device; and

generating, by the computing device, analytics for the campaign from the social signals of the second additional subset.

2. The method of claim 1, wherein identifying, by the computing device and at the first time, the first subset of the social signals comprises:

receiving second keywords, wherein the received second keywords are a priori associated with the campaign data, and wherein said first keyword is not a priori associated with the campaign at a time of the receipt of the second keywords; and

identifying the first subset using the received second keywords.

3. The method of claim 1, further comprising:

identifying a first number of ones of the social signals of said set that are associated with a brand;

identifying a second number of ones of social signals of said set that are associated with the campaign, wherein the campaign is associated with the brand; and

identifying a social media impact of the campaign based on the first number of the social signals that are associated with the brand and the second number of the social signals that are associated with the campaign.

4. The method of claim 1, further comprising:

identifying constituents associated with the campaign;

identifying a number of subscribers for the constituents; and

determining leveraged impressions for the constituents based on the number of subscribers for the constituents.

5. The method of claim 4, further comprising:

identifying a leveraged impression for a company based on a number of subscribers for accounts operated by the company; and

identifying a campaign lift based on a ratio between the leveraged impressions for the constituents and the leveraged impression for the company.

6. The method of claim 5, further comprising;

comparing the campaign lift with other campaign lifts for other campaigns to identify a relative success of the campaign.

7. The method of claim 1, further comprising identifying a campaign vocabulary for the campaign based on the known campaign terminology and the new campaign terminology.

8. The method of claim 7, wherein identifying the campaign vocabulary based on the known campaign terminology and the new campaign terminology comprises:

identifying a first group of the social signals of said collected set, the social signals of the first group associated with a brand for the campaign;

identifying a second group of the social signals of said collected set, the social signals of the second group associated with the campaign;

identifying a first set of frequently used terms in the social signals of the first group;

identifying a second set of frequently used terms in the social signals of the second group; and

identifying at least some of the second set of frequently used terms that are in the first set of frequently used terms as the campaign vocabulary.

9. The method of claim 8, further comprising:

identifying an advocate campaign vocabulary for advocates of the brand; and

identifying at least some of the advocate campaign vocabulary that is not contained in the campaign vocabulary as a positive campaign vocabulary.

10. The method of claim 9, further comprising:

identifying a detractor campaign vocabulary for detractors of the brand; and

identifying at least some of the detractor campaign vocabulary that is not contained in the campaign vocabulary as a negative campaign vocabulary.

11. The method of claim 1, further comprising:

identifying a first vocabulary from ones of the social signals of the second additional subset that are associated with a social group;

identifying a second vocabulary from ones of the social signals of the second additional subset that are associated with advocates of the brand; and

identifying at least some of the first vocabulary that is also contained in the second vocabulary as a positive brand vocabulary for the social group.

12. The method of claim 1, wherein the second additional subset of social signals is a superset with respect to the first subset of social signals.

13. The method of claim 1, wherein inputting the social signals of the first subset into the natural language processor to discover the new campaign terminology that is different than the known campaign terminology further comprises:

identifying a first group of the social signals of the first subset and a second different group of the social signals of the first subset;

inputting the social signals of the first group and the second group into the natural language processor to obtain first terms and second terms, respectively;

responsive to obtaining the first terms and the second terms from the natural language processor, filtering the first terms using the second terms; and

identifying said first keyword responsive to the filtering.

14. An apparatus, comprising:

circuitry configured to store a collected set of social signals generated on an electronic social network, the circuitry further configured to;

identify, at a first time, the social signals associated with a brand from the collected set of social signals generated on the electronic social network;

input the social signals associated with the brand into a natural language processor to discover new brand terminology that is different than known brand terminology for the brand, the new brand terminology not a priori associated with the brand by electronically stored data corresponding to the brand, wherein the new brand terminology is discovered at a second time that is later than the first time;

generate a brand vocabulary based on the known brand terminology and the new brand terminology;

identify a keyword based on the brand vocabulary;

identify, using the keyword, additional social signals associated with the brand vocabulary from the collected set of social signals generated on the electronic social network; and

generate analytics for the brand based on the identified social signals and the additional social signals associated with the brand.

15. The apparatus of claim 14, wherein the circuitry is further configured to:

identify a first group of ones of the social signals of said collected set that are associated with an ecosystem;

identify a first set of most frequently terms in the first group of social signals;

identify a second group of the social signals associated with the brand;

identify a second set of most frequently terms in the second group of social signals; and

generate the brand vocabulary based on the first set of most frequently used terms and the second set of most frequently used terms.

16. The apparatus of claim 15, wherein the circuitry is further configured to:

compare the first set of most frequently used terms with the second set of most frequently used terms; and

identify at least some of the second set of most frequently terms that are contained in the first set of most frequently used terms as the brand vocabulary.

17. The apparatus of claim 14, wherein the circuitry is further configured to:

identify a first group of the social signals associated with the brand;

identify a first set of most frequently terms in the first group of social signals;

identify a second group of ones of the social signals associated with a campaign conducted for the brand;

identify a second set of most frequently terms in the second group of social signals; and

generate a campaign vocabulary based on the first set of most frequently used terms and the second set of most frequently used terms.

18. The apparatus of claim 14, wherein the circuitry is further configured to:

identify a first group of ones of the social signals of the collected set that are associated with a user group;

identify a first set of most frequently terms in the first group of social signals;

identify a second group of ones of the social signals that are associated with an advocate of the brand;

identify a second set of most frequently terms in the second group of social signals; and

generate a positive brand vocabulary for the user group based on the first set of most frequently used terms and the second set of most frequently used terms.

19. The apparatus of claim 18, wherein the circuitry is further configured to:

identify a third group of ones of the social signals that are associated with detractor of the brand;

identify a third set of most frequently terms in the third group of social signals; and

identify a negative brand vocabulary for the user group based on the first set of most frequently used terms and the third set of most frequently used terms.

20. The apparatus of claim 14, wherein the circuitry is further configured to:

identify a first set of ones of the social signals associated with the brand that are generated during a first time period;

identify a second set of ones of the social signals associated with the brand that are generated during a second time period;

identify a first set of the terms used in the first set of the social signals;

identify a second set of the terms used in the second set of the social signals; and

detect a campaign associated with the brand based on the first set of the terms and the second set of the terms.

21. The apparatus of claim 20, wherein the circuitry is further configured to generate a campaign vocabulary for the campaign based on the first set of the terms and the second set of the terms.

22. A system, comprising:

a processing device configured to;

collect a set of social signals from different electronic social networks;

identify a first set of said collected set of social signals associated with a brand;

identify a second set of ones of said collected set of social signals that are associated with a campaign for the brand, the campaign defined by electronically stored campaign data, wherein the campaign data is associated with known campaign terminology and said identification of the second set of the social signals includes;

receive first keywords that are a priori associated with the campaign;

identify, at a first time, a grouping of the social signals of the collected set using the first keywords;

input the social signals of the grouping into a natural language processor to discover new campaign terminology that is different than the known campaign terminology, the new campaign terminology not a priori associated with the campaign by the campaign data, wherein the new campaign terminology is discovered at a second time that is later than the first time;

identify a second keyword based on the new campaign terminology; and

identify at least one of the social signals of said collected set that is not part of the grouping using the second keyword, wherein the second set of the social signals includes the social signals of the grouping and the at least one of the social signals of said collected set that is not part of the grouping; and

identify a first set of terms used in the first set of social signals;

identify a second set of terms used in the second set of social signals; and

generate a campaign vocabulary based on a comparison of the first set of terms with the second set of terms.

23. The apparatus of claim 22, wherein the processing device is further configured to:

identify a first set of most frequently used terms in the first set of terms;

identify a second set of most frequently used terms in the second set of terms; and

identify at least some of the second set of most frequently used terms that are not contained in the first set of most frequently used terms as part of the campaign vocabulary.

24. The apparatus of claim 23, wherein the processing device is further configured to:

identify a third set of ones of the social signals of the collected set that are associated with an advocate of the brand;

identify a third set of terms used in the third set of social signals;

generate a positive campaign vocabulary based on a comparison of the third set of terms with the campaign vocabulary.