From the Entity-Relationship Model to a Graph Database Model: Some Considerations and Challenges
Abstract
To meet the requirements of today’s big data applications, several types of databases have emerged in the last decade. One of them, the graph databases (GDB) focus on applications where a graph is an appropriate abstraction for the representation of the data and their relationships, e.g., a social network. The Model-Driven Design of Graph Databases methodology transforms an entity-relationship model into a conceptual model for a GDB. In this paper, we propose some ideas to address some problems of this methodology. In particular, we propose a hybrid approach, i.e., when in an application is only advisable to manage a part of the database through a GDB. We also used textual annotations from UML to specify some aspects of the resulting model and thus avoid some ambiguities that were not addressed in this methodology.