Workload-Aware Subgraph Query Caching and Processing in Large Graphs

A subgraph query q that finds as output all its subgraph-isomorphic embeddings from a data graph g has been core to modern declarative querying in large graphs. In this paper, we address subgraph queries with the availability of query workload information, W = {w
₁
,..., w
_n
}, where w
_i
∈ W is a previously issued query with all its subgraph-isomorphic embeddings cached beforehand. We introduce a workload-aware subgraph querying framework, WaSQ, that leverages query workload for subgraph query rewriting, search plan refinement, partial results reusing, and false positive filtering towards facilitating the whole subgraph querying process. Experimental studies in real-world graphs demonstrate that WaSQ achieves significant and consistent performance gains in comparison with state-of-the-art, workload-oblivious solutions for large-scale subgraph querying.