SQNR: A System for Querying Nodes and Relations in Multi-relational Social Networks

Social networks are usually modeled by graphs, where the nodes represent actors, and edges indicate interactions amongst actors. In the real world, social networks are mostly multi-relational, i.e., actors are related through different relationship types, such as friendship, kinship and co-authorship. A multi-relational social network can be represented by multiple single-relational graphs, each reflecting one kind of relation. Acquiring information contained in a multi-relational social network, such as which actors interact frequently in which relationships, can help us to understand the network comprehensively and then offer better services. This paper develops a system (SQNR: A System for Querying Nodes and Relations) to query nodes that interact frequently in the specified relationships and to query relationships in which the specified community patterns (limitations of the community memberships of some actors, such as belong to the same community) can be satisfied. Twelve community patterns can be specified in this paper. The queried results provide new insights about relationships amongst actors.