A Selection Process of Graph Databases Based on Business Requirements

Abstract

Several graph databases provide support to analyze a large amount of highly connected data, and it is not trivial for a company to choose the right one. We propose a new process that allows analysts to select the database that suits best to the business requirements. The proposed selection process makes possible to benchmark several graph databases according to the user needs by considering metrics such as querying capabilities, built-in functions, performance analysis, and user experience. We have selected some of the most popular native graph database engines to test our approach to solve a given problem. Our proposed selection process has been useful to design benchmarks and provides valuable information to decide which graph database to choose. The presented approach can be easily applied to a wide number of applications such as social network, market basket analysis, fraud detection, and others.