Iceberg Clique queries in large graphs

초록

This paper investigates the Iceberg Clique (IC) queries in a large graph, specially, given a user-specified threshold theta, an IC query reports the cliques where the number of vertices exceeds left perpendicular theta|V| right perpendicular. Toward this end, a practical IC query theorem is formally proposed and proved. With this proposed query theorem, a formal context and its corresponding iceberg concept lattice are first constructed from an input graph topology by Modified Adjacency Matrix; then, we prove that the IC queries problem is equivalent to finding the iceberg equiconcepts whose number of elements exceeds left perpendicular theta|V| right perpendicular. Theoretical analysis and experimental results demonstrate that the proposed query algorithm is feasible and efficient for finding the iceberg cliques from large graphs. (C) 2017 Published by Elsevier B.V.

키워드