Apparatus and method for model-based social analytics

• US 9,251,530 B1
• Filed: 08/31/2012
• Issued: 02/02/2016
• Est. Priority Date: 08/31/2012
• Status: Active Grant

• Current Assignee: TPG Specialty Lending Incorporated
• Original Assignee: Sprinklr Inc.
• Inventors: Huddleston, Erik Lee, Dachis, Jeffrey, Huddleston, Brian Justin II, Vaughan, Steven Michael, Cauvin, Roger Louis, De Oliveira, John Joseph, Eason, William Lance, Hanna, Jeremy, Hardt, Shannon Paul, Horne, Bryan Joseph, Kearby, Brandon, Miller, Clinton Frederick, Perkins, Jacob Andrew, Potter, Timothy Joseph, Renteria, Ramon, Westigard, Jason Samuel, Yu, David Chi-Fine
• Primary Examiner(s): Etienne, Ario
• Assistant Examiner(s): SHIU, HO T
• Application Number:
  US13/601,151
• Time in Patent Office: 1,250 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0242 Determining effectiveness of advertisements
• CPC Class Codes: G06Q 50/01 Social networking

Abstract

A model based social analytic system collects social signals from social network accounts for different companies and constituents. The social signals can be used to benchmark social network performance for different contextual dimensions, such as industries, companies, brands, etc. Conversations may be identified in the signals and the types of constituents participating in the conversations may be identified. Analytics can then be determined for the contextual dimensions based on the related conversations and the types of constituents participating in the conversations.

Claims

1. A method, comprising:

identifying accounts associated with social networks;

collecting signals for the accounts;

associating contextual dimensions with the accounts;

identifying relationships between the accounts and/or signals and the contextual dimensions;

generating analytics for the contextual dimensions based on the relationships;

identifying the signals containing messages for a conversation discussing a same topic;

identifying a first type of user participating in the conversation as influencers that have a plurality of followers;

identifying a second type of user participating in the conversation as advocates;

determining a first weighted score based at least in part on applying a first weight to a first portion of the messages generated by the influencers;

determining a second weighted score based at least in part on applying a second weight to a second portion of the messages generated by the advocates;

calculating a strength of the conversation based at least in part on the first weighted score and the second weighted score;

generating a score for the topic based at least in part on the strength of the conversation; and

displaying the score of the topic on a display.

2. The method of claim 1, further comprising identifying a relevance of the signals or accounts to the contextual dimensions and generating the analytics for the contextual dimensions based on the relevance.

3. The method of claim 1, further comprising:

periodically generating the analytics for the contextual dimensions for different time periods;

using a first set of the analytics for a first one of the time periods as a benchmark; and

comparing the first set of analytics to a second set of the analytics for a second one of the time periods to identify changes in the analytics.

4. The method of claim 1, comprising:

identifying a first set of the signals associated with a first one of the contextual dimensions;

generating a first set of analytics for the first one of the contextual dimensions from the first set of the signals;

identifying a second set of the signals associated with a second one of the contextual dimensions;

generating a second set of analytics for the second one of the contextual dimensions from the second set of the signals;

using the first set of analytics as a benchmark for comparing with the second set of analytics.

5. The method of claim 4, wherein the first one of the contextual dimensions comprises an industry and the second one of the contextual dimensions comprises a brand within the industry.

6. The method of claim 4, wherein the first one of the contextual dimensions comprises a brand and the second one of the contextual dimensions comprises a competitor brand.

7. The method of claim 4, wherein the first one of the contextual dimensions comprises a company and the second one of the contextual dimensions comprises a brand sold by the company.

8. The method of claim 4, wherein the first one of the contextual dimensions comprises a brand and the second one of the contextual dimensions comprises a geographic region for the brand.

9. The method of claim 1, further comprising:

identifying the accounts associated with the contextual dimensions as constituents of the contextual dimensions;

identifying the signals associated with the constituents; and

generating the analytics for the contextual dimensions based on the signals associated with the constituents of the contextual dimensions.

10. The method of claim 9, further comprising:

identifying the constituents generating positive messages about the associated contextual dimensions as advocates of the contextual dimensions;

identifying the constituents generating negative messages about the associated contextual dimensions as detractors of the contextual dimensions; and

identifying the constituents associated with the contextual dimensions and having a threshold number of followers as influencers of the contextual dimensions.

11. The method of claim 1, further comprising:

identifying the accounts associated with companies as company accounts;

identifying the accounts for employees of the companies as employee accounts; and

generating the analytics for the company accounts based on the signals associated with the employee accounts.

12. The method of claim 1, further comprising:

identifying social network sources associated with the contextual dimensions and having a threshold number of followers as influencers of the contextual dimensions;

adding the social network sources to the accounts; and

collecting signals from the social network sources added to the accounts.

13. The method of claim 1 further comprising:

calculating the strength of the conversation based at least in part on a sentiment of the messages for the conversation, wherein the sentiment includes positive, neutral, and negative reviews in the messages discussing the topic.

14. An analytic system, comprising:

a computing system including a processor configured to execute software instructions stored in memory, the software instructions comprising;

a collector module configured to collect signals from sources in a social network, wherein the sources are identified by accounts;

an enrichment module configured to identify;

relationships between the signals and the accounts;

the signals containing messages for a conversation discussing a same topic;

a first type of user participating in the conversation as influencers that have a plurality of followers; and

a second type of user participating in the conversation as advocates;

an analysis module configured to;

determine a first weighted score based at least in part on applying a first weight to a first portion of the messages generated by the influencers;

determine a second weighted score based at least in part on applying a second weight to a second portion of the messages generated by the advocates;

calculate a strength of the conversation based at least in part on the first weighted score and the second weighted score;

generate a score for the topic based at least in part on the strength of the conversation; and

a display module configured to display the score of the topic on a display device.

15. The analytic system of claim 14, further comprising multiple collector modules configured to collect the signals from different social networks and store the collected signals into time buckets, wherein some of the signals are collected by the collector modules polling the social networks and some of the signals are collected by the social networks streaming the signals to the collector modules.

16. The analytic system of claim 14, further comprising a normalize module configured to load data from the signals into a table and add identifiers to the signals identifying the accounts associated with the signals and identifying ecosystems for the accounts.

17. The analytic system of claim 14, wherein the enrichment module is further configured to identify contextual dimensions associated with the accounts.

18. The analytic system of claim 17, wherein the contextual dimensions comprise at least one of an industry associated with the accounts, a company associated with the accounts, brands or services associated with the accounts, and/or geographic regions associated with the accounts.

19. The analytic system of claim 17, wherein the enrichment module is further configured to identify some of the accounts as constituent accounts for other related accounts.

20. The analytic system of claim 19 wherein the constituent accounts comprise at least one of:

advocate accounts generating positive messages for the related accounts;

employee accounts for employees of companies associated with the related accounts;

partner accounts for the accounts of business partners of the companies associated with the related accounts;

market accounts having some interactions with the related accounts; and

/or

influencer accounts having some interactions with the related accounts and having a threshold number of followers or subscribers.

21. The analytic system of claim 14, wherein the enrichment module is further configured to identify the signals associated with same conversations.

22. The analytic system of claim 21, wherein the enrichment module is further configured to:

identify the signals initiating the conversations;

identify types of messages used in the signals associated with the conversations;

identify contextual dimensions associated with the conversations; and

identifying a sentiment for the conversations.

23. The analytic system of claim 22, wherein the contextual dimensions comprise at least one of an industry, a company, a brand, and/or a geographic region associated with the conversation.

24. The analytic system of claim 14, wherein the analysis module is further configured to:

identify a social network source generating some of the signals;

identify the social network source has having a large number of subscribers; and

direct the collector module to add the social network source to the accounts and start collecting signals from the social network source.