Cohesive Subgraph Detection in Large Bipartite Networks

Cohesive Subgraph Detection in Large Bipartite Networks

연구 분야: Safety

학회: SSDBM '20: Proceedings of the 32nd International Conference on Scientific and Statistical Database Management

In real-world applications, bipartite graphs are widely used to model the relationships between two types of entities, such as customer-product relationship, gene co-expression, etc. As a fundamental problem, cohesive subgraph detection is of great importance for bipartite graph analysis. In this paper, we propose a novel cohesive subgraph model, named (α, β, ω)-core, which requires each node should have sufficient number of close neighbors. The model emphasizes both the engagement of entities and the strength of connections. To scale for large networks, efficient algorithm is developed to compute the (α, β, ω)-core. Compared with the existing cohesive subgraph models, we conduct the experiments over real-world bipartite graphs to verify the advantages of proposed model and techniques.