Attributed Community Search with Forbidden Nodes

The study of community search problems in large graphs has attracted more and more attention: given a query node q, it is required to return communities containing the query node. In recent years, some researchers have considered keyword information to make the attributed community search is gradually emerging, which is widely used in social network marketing and setting up of social events. However, some problems require that the query communities cannot contain certain forbidden nodes, which has not received any attention in previous work. Based on this, this paper proposes undirected attributed community search with forbidden nodes problem (FATQ), which not only avoids forbidden nodes from existing in the query communities, but also guarantees the structure cohesiveness of the query communities and the keyword cohesiveness of the communities. Finally, this paper proposes two different search algorithms for FATQ problem and evaluates our algorithms on multiple datasets. The results show that correctness and effectiveness of the proposed methods.