A Novel Subgraph Querying Method on Directed Weighted Graphs

The usage of graphs has led to the emergence of schema queries in knowledge graph and graph databases, where subgraph queries have become one of the most important research problems. In this paper, we study the directed weighted graphs, and propose a subgraph querying method NGraph based on shortest weight paths. Specifically, we extract three features: vertices, edges and shortest weight paths, which can effectively describe a data graph. Then, the extracted three features are encoded according to corresponding coding approaches and the coding results are combined to form vertex codes and graph codes. The index tree is then built by the encoding of the graph, which is each of the graph sets. According to the filtering-and-verification framework, the candidate set is obtained. Finally, the result set is verified according to VF2 algorithm. The experimental results show that the proposed method can accelerate the querying on directed graphs, and thus improves the querying efficiency.