Bivariate, cluster, and suitability analysis of NoSQL solutions for big graph applications

With the explosion of social media, the Web, Internet of Things, and the proliferation of smart devices, large amounts of data are being generated each day. However, traditional data management technologies are increasingly inadequate to cope with this growth in data. NoSQL has become increasingly popular as this technology can provide consistent, scalable and available solutions for the ever-growing heterogeneous data. Recent years have seen growing applications shifting from traditional data management systems to NoSQL solutions. However, there is limited in-depth literature reporting on NoSQL storage technologies for big graph and their applications in various fields. This chapter fills this gap by conducting a comprehensive study of 80 state-of-the-art NoSQL technologies. In this chapter, we first present a feature analysis of the NoSQL solutions and then generate a data set of the investigated solutions for further analysis in order to better understand and select the technologies. We perform a clustering analysis to segment the NoSQL solutions, compare the classified solutions based on their storage data models and Brewer's CAP theorem, and examine big graph applications in six specific domains. To help users select appropriate NoSQL solutions, we have developed a decision tree model and a web-based user interface to facilitate this process. In addition, the significance, challenges, applications and categories of storage technologies are discussed as well.