Two-Level Index for Truss Community Query in Large-Scale Graphs
Zheng Lu, Yunhe Feng, Qing Cao · 2019
Recently, there has been a significant interest in the study of the community search problem in large- scale graphs. K-truss as a community model has drawn increasing attention in the literature. In this work, we extend our scope from the community search problems to a more generalized local community query problem based on a triangle- connected k-truss community model. We classify local community query into two categories, community-level and edge-level query, based on the information required to process a given query. We design a two-level index structure that supports both types of queries with multiple query vertices and arbitrary cohesiveness criteria. We conduct extensive experiments using real-world large-scale graphs and compare with the state-of-the-art methods of k-truss community search. The results show that our method outperforms the state-of-the- art works in various types of local k-truss community queries.