Open AccessA Rule-based Conversion of an EER Schema to Neo4j Schema Constraints
Author(s) -
Telmo Henrique Valverde da Silva,
Ronaldo dos Santos Mello
Publication year - 2021
Language(s) - English
Resource type - Conference proceedings
DOI - 10.5753/sbbd.2021.17876
Subject(s) - nosql , computer science , schema (genetic algorithms) , conceptual schema , database schema , schema evolution , semi structured model , star schema , data mining , information retrieval , database , big data , database design , developmental psychology , gender schema theory , psychology
Several application domains hold highly connected data, like supply chain and social network. In this context, NoSQL graph databases raise as a promising solution since relationships are first class citizens in their data model. Nevertheless, a traditional database design methodology initially defines a conceptual schema of the domain data, and the Enhanced Entity-Relationship (EER) model is a common tool. This paper presents a rule-based conversion process from an EER schema to Neo4j schema constraints, as Neo4j is the most representative NoSQL graph database management system with an expressive data model. Different from related work, our conversion process deals with all EER model concepts and generates rules for ensuring schema constraints through a set of Cypher instructions ready to run into a Neo4j database instance, as Neo4J is a schemaless system, and it is not possible to create a schema a priori. We also present an experimental evaluation that demonstrates the viability of our process in terms of performance.