Label inference in a social network

Abstract

At least one embodiment of this disclosure includes a method of inferring attribute labels for a user in a social networking system based on the user's social connections and user-specified attribute labels in the social networking system. The method can include: establishing variational equations based on attribute labels of nodes in an ego network in a social graph of a social networking system; determining likelihood scores for at least a portion of the attribute labels of neighboring nodes from a focal user node in the ego network based on user-specified attribute labels from the social networking system; and calculating probability distributions of possible attribute labels for the focal user node of the ego network based on the variational equations and the likelihood scores.

1 - 15 . (canceled) 
     
     
         16 . A computer-implemented method, comprising:
 establishing variational equations based on attribute labels of nodes in an ego network in a social graph of a social networking system;   determining likelihood scores for at least a portion of the attribute labels of neighboring nodes from a focal user node in the ego network based on user-specified attribute labels from the social networking system;   computing probability distributions of possible attribute labels for the focal user node of the ego network based on the variational equations and the likelihood scores; and   providing at least a portion of the computed probability distributions with an external process instance inferring labels for another ego network of the social networking system.   
     
     
         17 . The computer-implemented method of  claim 16 , further comprising:
 receiving an update to a first attribute label of at least one of the neighboring nodes, wherein the update is a user-reported attribute label or a probability distribution update of possible attribute labels from the external process instance inferring labels for another ego network of the social networking system; and   re-calculating re-computing the probability distributions based at least partly on the update.   
     
     
         18 . The computer-implemented method of  claim 16 , further comprising:
 inferring an attribute label based on the probability distributions for the focal user node; and   generating a user prompt based on the inferred attribute label for advertisement, friend suggestion, label suggestion, or affiliation suggestion.   
     
     
         19 . The computer-readable storage medium of  claim 16 , wherein computing the probability distribution includes inferring an explanation for a formation of a social connection in the ego network. 
     
     
         20 . A computer system, comprising:
 a component configured to establish variational equations based on attribute labels of nodes in a first ego network in a social graph of a social networking system;   a component configured to set likelihood scores for at least a portion of the attribute labels of neighboring nodes from a focal user node in the ego network based on user-specified attribute labels from the social networking system;   a component configured to compute probability distributions of possible attribute labels for the focal user node of the ego network based on the variational equations and the likelihood scores; and   a component configured to share at least a portion of the computed probability distributions with another process instance of a method inferring labels for a second ego network of the social networking system.   
     
     
         21 . The system of  claim 20 , further comprising:
 a component configured to receive an update to a first attribute label of at least one of the neighboring nodes, wherein the update is a user-reported attribute label or a probability distribution update of possible attribute labels from an external process instance configured to infer labels for another ego network of the social networking system; and   a component configured to re-compute the probability distributions based at least partly on the update.   
     
     
         22 . The system of  claim 20 , further comprising:
 a component configured to infer an attribute label based on the probability distributions for the focal user node; and   a component configured to generate a user prompt based on the inferred attribute label for advertisement, friend suggestion, label suggestion, or affiliation suggestion.   
     
     
         23 . The system of  claim 20 , wherein computing the probability distribution includes inferring an explanation for a formation of a social connection in the ego network. 
     
     
         24 . A non-transitory computer-readable medium storing instructions, the instructions comprising:
 instructions for establishing variational equations based on attribute labels of nodes in an ego network in a social graph of a social networking system;   instructions for determining likelihood scores for at least a portion of the attribute labels of neighboring nodes from a focal user node in the ego network based on user-specified attribute labels from the social networking system;   instructions for computing probability distributions of possible attribute labels for the focal user node of the ego network based on the variational equations and the likelihood scores; and   instructions for providing at least a portion of the computed probability distributions with an external process instance inferring labels for another ego network of the social networking system.   
     
     
         25 . The computer-implemented method of  claim 24 , further comprising:
 instructions for receiving an update to a first attribute label of at least one of the neighboring nodes, wherein the update is a user-reported attribute label or a probability distribution update of possible attribute labels from the external process instance inferring labels for another ego network of the social networking system; and   instructions for re-computing the probability distributions based at least partly on the update.   
     
     
         26 . The computer-implemented method of  claim 24 , further comprising:
 instructions for inferring an attribute label based on the probability distributions for the focal user node; and   instructions for generating a user prompt based on the inferred attribute label for advertisement, friend suggestion, label suggestion, or affiliation suggestion.   
     
     
         27 . The computer-readable storage medium of  claim 24 , wherein the instructions for computing the probability distribution include instructions for inferring an explanation for a formation of a social connection in the ego network.