Efficient access methods for very large distributed graph databases

• Indexing techniques are essential in large scale subgraph searching. Three new indexing proposed, which leverage the use of bitmaps. Generic framework for filter-then-verify implementations on top Apache Spark. Evaluation shows that different indexes suitable query selectivities. A distributed approach is very databases and low selective queries. Subgraph searching an problem graph databases, but it also challenging due to involved isomorphism NP-Complete sub-problem. Filter-Then-Verify (FTV) methods mitigate performance overheads by using index prune out graphs do not fit a filtering stage, reducing number evaluations subsequent verification stage. has be applied (tens millions graphs) real applications such as molecular substructure Previous surveys have identified FTV solutions GraphGrepSX (GGSX) CT-Index best ones (thousands graphs), however they cannot reach reasonable graphs). This paper proposes generic implementation solutions. Besides, three previous improve GGSX adapted executed clusters. The evaluation how achieved provide great improvement (between 70% 90% time reduction) centralized configuration may used achieve efficient over cluster configurations.