Model-Driven Engineering of an OpenCypher Engine: Using Graph Queries to Compile Graph Queries

Abstract

Abstract. Graph database systems are increasingly adapted for storing and processing heterogeneous network-like datasets. Many challenging applications with near real-time requirements-such as financial fraud detection, on-the-fly model validation and root cause analysis-can be formalised as graph problems and tackled with graph databases efficiently. However, as no standard graph query language has yet emerged, users are subjected to the possibility of vendor lock-in. The openCypher group aims to define an open specification for a declarative graph query language. However, creating an openCypher-compatible query engine requires significant research and engineering efforts. Meanwhile, model-driven language workbenches support the creation of domainspecific languages by providing high-level tools to create parsers, editors and compilers. In this paper, we present an approach to build a compiler and optimizer for openCypher using model-driven technologies, which allows developers to define declarative optimization rules.

Cypher

“…The language therefore is used in hundreds of production applications across many industry vertical domains. Cypher is also used in several research projects (e.g., Ingraph [41], Gradoop [29], and Cytosm [55]) as well as in recent or incubating open-source projects, such as Cypher for Apache Spark and Cypher over Gremlin.…”