Cohesive Subgraph Identification in Weighted Bipartite Graphs

Cohesive subgraph identification is a fundamental problem in bipartite graph analysis. In real applications, to better represent the co-relationship between entities, edges are usually associated with weights or frequencies, which neglected by most existing research. To fill gap, we propose new cohesive model, (k,ω)-core, considering both cohesiveness and frequency for weighted graphs. Specifically, (k,ω)-core requires each node on left layer have at least k neighbors (cohesiveness) right weight of ω (frequency). scenarios, different users may parameter requirements. handle massive graphs queries, index-based strategies developed. addition, effective optimization techniques proposed improve index construction phase. Compared baseline, extensive experiments six datasets validate superiority our methods.