Methods and systems for inferring user attributes in a social networking system

A set of one or more computer-readable non-transitory storage media that provides instructions that, when executed by a set of one or more processing devices, will cause said set of processing devices to perform operations comprising:

• maintaining one or more data stores storing a social graph of a social networking system comprising a plurality of nodes and a plurality of edges between the nodes, wherein the plurality of nodes includes user nodes corresponding to users of the social networking system and concept nodes corresponding to concepts;
• identifying an unknown, incomplete, or inaccurate user attribute to be inferred for a user of the social networking system;
• generating a plurality of probability lists using a corresponding plurality of probability algorithms that utilize a set of known user attributes of the user and the social graph, wherein each probability list of the plurality of probability lists includes a set of one or more probability entries that each include a prediction value and a confidence score corresponding to the prediction value, wherein the prediction value is a possible value of the unknown, incomplete, or inaccurate user attribute, and wherein the confidence score is a value indicating a predicted likelihood that the prediction value is a correct value of the unknown, incomplete, or inaccurate user attribute;
• generating an inferred user attribute value based upon the plurality of probability lists and a plurality of weights corresponding to the plurality of probability algorithms, wherein each weight of the plurality of weights indicates a relative confidence that the corresponding probability algorithm will generate a probability list including a prediction value that is the correct value of the unknown, incomplete, or inaccurate user attribute;
• and
• storing the inferred user attribute value in the one or more data stores.