Social content connection system and method

A computer-implemented method of recommending content to users of a social network based on values embodied in a social graph the improvement comprising:

• monitoring and measuring activities of a set of users with a computing system at a social networking site to create explicit edges and nodes for the social graph connecting users to each other as well as to a set of items based on ratings and friend designations made by said set of users;
• wherein the social graph is stored in an electronic data file identifying said edges, said nodes and associated values of each;
• analyzing said social graph to identify data sparsities corresponding to a lack of explicit ratings data for said set of users and/or said set of items;
• selecting one of a first user and a second user connected in said social graph to said first user from said set of users based on identifying said data sparsities;
• soliciting a first prediction with the computing system from said second user concerning a first rating that said first user would give to a target item in said set of items;
• wherein said target item is represented by a first node that is not yet connected with an explicit edge to said first user in said social graph;
• creating at least a tentative edge with the computing system in said social graph connecting said first node for said target item to said first user based in part at least on said first prediction;
• wherein recommendations for one of said set of items, including said target item, are generated by a recommender computing system for said first user based in part at least on both explicit edges and tentative edges in said social graph.