MHNA: Multi-Hop Neighbors Aware Index for Accelerating Subgraph Matching

With the proliferation of knowledge graphs in various domains, efficient processing of subgraph matching queries has become a crucial issue. However, the subgraph matching problem has been proven to be an NP-complete problem. While specific approaches aim to accelerate queries by leveraging favorable matching orders and pruning rules, they face limitations in handling large-scale graph data due to the exponential search space. Conversely, other methods employ graph indexes to enhance query efficiency, but these indexes provide limited acceleration capabilities or encounter challenges in widespread adoption due to their substantial size. In this paper, a novel index called MHNA (Multi-Hop Neighbors Aware) is proposed, which is devised to accelerate subgraph matching while minimizing the space overhead. Moreover, we introduce an efficient iterative index construction method that computes the MHNA index for each vertex by solely leveraging its neighboring information. Extensive experiments demonstrate that our methods reduce storage space and construction time by an order of magnitude compared to existing the state-of-the-art database indexing approaches. Regarding query processing time, MHNA achieves up to a 30-fold reduction in time overhead.