Automated generation of consistent, diverse and structurally realistic graph models

Abstract In this paper, we present a novel technique to automatically synthesize consistent, diverse and structurally realistic domain-specific graph models. A model is (1) consistent if it metamodel-compliant satisfies the well-formedness constraints of domain; (2) local neighborhoods nodes are highly different; synthetic at close distance representative real according various metrics used in network science, databases or software engineering. Our approach grows models by extension operators using hill-climbing strategy way that (A) ensures there no constraint violation (for consistency reasons), while (B) more candidates selected minimize target metric value (wrt. model). We evaluate effectiveness for generating multiple guidance heuristics compared other generators context three case studies with large set human also highlight our able generate models, which requirement many testing scenarios.