Keyword-Constrained Community Search over Bipartite Networks

Bipartite networks are widely used to model interactions between two types of entities in many real-world applications, such as recommendation systems, social networks, and academic networks. Community search aims to identify cohesive groups containing query entities, while keyword-constrained community search further incorporates semantic requirements. However, keyword-constrained community search over large-scale unweighted bipartite graphs remains insufficiently explored. In this paper, we investigate the keyword-constrained bi-community search problem over unweighted bipartite graphs and propose an efficient graph analytics framework. Specifically, we develop two search algorithms based on global and local search paradigms as baseline solutions, and further propose a Hybrid Index (HI) that integrates keyword inverted lists with structural core-based labels to accelerate query processing. Extensive experiments on four real-world bipartite graph datasets demonstrate that the proposed HI-based approach achieves significant improvements in query efficiency compared with the baseline algorithms.