Data Modeling for Connected Data -- A systematic literature review

TL;DR

This paper systematically reviews approaches to data modeling for connected data in NoSQL graph databases, highlighting the challenges and research gaps, especially in reverse engineering methods, over the period 2013-2020.

Contribution

It provides a comprehensive analysis of sixteen studies on connected data modeling, identifying research opportunities and characterizing existing approaches.

Findings

Reverse engineering of connected data is underexplored.

Most works focus on modeling techniques for specific applications.

There is a need for more complete solutions in data reverse engineering.

Abstract

A data model specifies how real-world entities and their relationships are represented and operated. In the NoSQL world data modeling usually begins from identifying application queries and designing the data model to efficiently answer them so each database is designed to meet requirements of just one or more applications. But this practice causes a strong coupling between the data model and application queries and promotes data silos. Newly developed applications that manipulate connected data, usually stored in NoSQL Graph Databases, suffer from this type of problem, which is a challenge for data integration projects in Big Data scenarios. This systematic literature review (SLR) was carried out to identify the known approaches for data modeling of connected data. The main contribution of this SLR is an analysis of sixteen works, from 2013 to 2020, in terms of three dimensions: type of contribution, bibliometrics, and data modeling characteristics. Through this analysis, it was possible to identify that reverse engineering of connected data is still a research opportunity since few and incomplete works were found.