Introduction

Abstract

Graph data management systems have experienced a renaissance in recent years. The reason for this is clear: with a confluence of trends in society, science, and technology, graph-structured data sets are increasingly being constructed, collected, and made available for analysis. Common everyday examples of massive and ever-growing graph data collections include social, biochemical, ecological, citation, communication, mobility, and transportation networks [Newman, 2018]. Graph data modeling and querying arises in such applications where the primary focus is on things and their relationships and the rich patterns in these complex webs of connectivity. In a social network such as LinkedIn¹ or Viadeo,² for example, we primarily have people and institutions as the things (i.e., nodes, vertices) and social connections such as “follows” and “works for” as the relationships (i.e., edges). In this domain, a job-seeker may be interested in answers to queries such as “Who are the people in my social network with a shared professional society membership who live in my city?” As another example, in biological network data sources such as BioGRID³ or UniProtKB,⁴ we typically have entities such as proteins as the things and the interactions between proteins as the relationships. Here, a scientist might be interested in querying for interaction pathways which have not been explored before in the literature, toward insight for new medical treatments. Teasing out such hidden patterns in graph databases are often a basis for knowledge and value creation in many contemporary application domains across the sciences and society.