Determining brand exclusiveness of users

Abstract

Embodiments of the present invention relate to a determination of a user's exclusiveness toward a particular brand. User-specific entities are extracted from social media content associated with a user. At least a portion of the user-specific entities are brand-related entities that are specifically relevant to a particular brand. These brand-related entities are analyzed with respect to the user-specific entities extracted from the social media content to determine a level of exclusivity of the user to the brand.

What is claimed is: 
     
         1 . One or more computer storage media storing computer-useable instructions that, when used by one or more computing devices, cause the one or more computing devices to perform operations comprising:
 generating, by a brand entity module of a computing device, a weighted distribution of brand-related entities that are associated with a particular brand indicating an importance of each brand-related entity with respect to a set of brand-specified entities;   extracting for each user of a plurality of users, by a user entity module of the computing device, a plurality of user entities relevant to the particular brand from social media content corresponding to each user;   based at least on a determined overlap between the weighted distribution of brand-related entities and a weighted distribution of user entities, determining, by a scoring module of the computing device, a level of exclusivity of each user to the brand; and   filtering, by the computing device, the plurality of users to generate a set of users that is exclusive to the brand based on the level of exclusivity of each user to the brand.   
     
     
         2 . The one or more computer storage media of  claim 1 , further comprising identifying the set of brand-specified entities by analyzing a plurality of data associated with the particular brand, wherein the set of brand-specified entities are of interest to a company associated with the particular brand. 
     
     
         3 . The one or more computer storage media of  claim 2 , further comprising generating a frequency distribution of the entities in the set of brand-specified entities. 
     
     
         4 . The one or more computer storage media of  claim 3 , wherein the set of brand-specified entities comprises a first set of entities that is identified from the plurality of data associated with the brand, and a second set of entities having a high co-occurrence with the entities in the first set of entities. 
     
     
         5 . The one or more computer storage media of  claim 1 , further comprising generating a frequency distribution for the plurality of user entities extracted from the social media content. 
     
     
         6 . The one or more computer storage media of  claim 5 , wherein at least a portion of the plurality of entities extracted from the social media content are not relevant to the particular brand, and further comprising applying a thresholding function to the frequency distribution. 
     
     
         7 . The one or more computer storage media of  claim 1 , further comprising calculating, for each entity in the weighted distribution of user entities, a term-interest indicating a tendency of exclusivity based on a frequency of each entity with respect to the plurality of user entities. 
     
     
         8 . The one or more computer storage media of  claim 1 , wherein the overlap is determined by:
 calculating a proportion of the plurality of user entities that are in the weighted distribution of brand-related entities; and   calculating a proportion of the brand-related entities that are in the weighted distribution of user entities.   
     
     
         9 . The one or more computer storage media of  claim 8 , further comprising:
 determining a weight for each entity of the brand-related entities with respect to the weighted distribution of brand-related entities; and   determining a weight for each user entity with respect to the weighed distribution of user entities.   
     
     
         10 . The one or more computer storage media of  claim 9 , further comprising:
 based on the weighted distribution of brand-related entities, providing a ranking for at least a portion of the brand-related entities;   based on the weighted distribution of user entities, providing a ranking for at least a portion of the user entities; and   calculating a rank distance based on a difference between the rankings for corresponding entities in the brand-related entities and the user entities.   
     
     
         11 . The one or more computer storage media of  claim 10 , further comprising:
 calculating a score indicating an exclusiveness of each user to the brand based at least on the rank distance.   
     
     
         12 . The one or more computer storage media of  claim 1 , wherein the social media content is extracted from one or more social networking services with which the user has interacted. 
     
     
         13 . A computer-implemented method comprising:
 generating, by a brand entity module of a computing device, a distribution of a first set of entities, the first set of entities extracted from data obtained from a company associated with a particular brand;   extracting, by a user entity module of the computing device, a second set of entities from social media content associated with one or more social networking services with which a user has interacted;   generating, by the user entity module of the computing device, a weighted distribution of the second set of entities, wherein each entity in the second set of entities is weighted based on a total number of entities mentioned by the user in the social media content;   ranking each entity in the second set of entities with respect to the weighted distribution of the second set of entities;   calculating a rank distance based on a difference between the rankings for corresponding entities in the first set of entities and the second set of entities;   based at least on the rank distance and a determined overlap between the first set of entities and the second set of entities, generating a score, by a scoring module of the computing device, indicating a level of exclusivity of the user to the particular brand; and   based on the score, determining whether a user belongs to a set of users that is exclusive to the particular brand.   
     
     
         14 . The computer-implemented method of  claim 13 , wherein the distribution of the first set of entities is a weighted distribution of each entity in the first set of entities with respect to the totality of the first set of entities. 
     
     
         15 . The computer-implemented method of  claim 13 , further comprising:
 generating a frequency distribution of the second set of entities extracted from social media content; and   applying a thresholding function to at least one of the frequency distribution and the weighted distribution of the second set of entities.   
     
     
         16 . The computer-implemented method of  claim 13 , wherein the overlap is determined by calculating a proportion of the first set of entities in the weighted distribution of the second set of entities and calculating proportion of the second set of entities in the distribution of the first set of entities. 
     
     
         17 . The computer-implemented method of  claim 14 , further comprising:
 determining a weight for each entity of the first set of entities with respect to the weighed distribution of the first set of entities; and   determining a weight for each entity in the second set of entities with respect to the weighted distribution of the second set of entities.   
     
     
         18 . The computer-implemented method of  claim 13 , wherein the first set of entities comprises a set of correlated entities that have a high co-occurrence or correlation value with entities specified by the company associated with the particular brand. 
     
     
         19 . A computerized system comprising:
 at least one processor; and   at least one computer storage medium storing computer useable instructions that when used by the at least one processor, cause the at least one processor to:   generate, by a brand entity component of a computing device, a weighted distribution of brand-related entities that are associated with a particular brand, the weighted distribution indicating an importance of each brand related entity with respect to the totality of brand-related entities;   identify, by the computing device, a plurality of users who have interacted with at least one social networking service, each user having corresponding social media content;   extract, for each user of the plurality of users, by a user entity module of the computing device, a plurality of user-specific entities from the social media content, wherein the plurality of user-specific entities are relevant to the particular brand;   generate, by the user entity module of the computing device, a weighted distribution of user-specific entities;   calculate rankings for the brand-related entities and the user-specific entities, wherein the rankings are based on the weighted distribution of brand-related entities and the weighted distribution of user-specific entities;   utilizing the rankings as inputs to a distance function, calculate a rank distance between corresponding entities in the brand-related entities and the user-specific entities;   based, at least, on the rank distance and a determined overlap between the brand-related entities and the user-specific entities, generating, by a scoring module of the computing device, a score for each user of the plurality of users indicating a level of exclusivity of a user to the particular brand; and   filtering, by the computing device, the plurality of users to generate a set of users that is exclusive to the particular brand based on the score generated for each user.   
     
     
         20 . The computerized system of  claim 19 , further comprising identifying the brand-related entities based on at least one of: an analysis of company data associated with the particular brand, wherein the company data is from the at least one social networking service, an analysis of a dataset provided by a company associated with the particular brand, and entities explicitly defined by a company associated with the particular brand.