Recent interest in graph databases has led to questions on selecting the appropriate graph database or graph analysis tool for a given solution. In this tutorial, we will introduce attendees to the two graph data models: the Resource Description Framework (RDF) model and the Property Graph model. RDF and property graph are both flexible, well-known data models suitable for modeling entities and relationships. Since they are both fundamentally graph models, they exhibit similarities, but also have some striking differences in their feature set and intended uses. For example, RDF, an open W3C standards graph with built-in uniqueness for every node and edge, is optimized for semantic searching and for supporting a unified metadata layer across one or more graphs, disparate datasets, and linked data. On the other hand, a property graph is node-centric, whereby relationships, represented as edges, connect the nodes to form the structure of the graph. Being node-centric enables social network analysis to detect communities and components, evaluate community structures, rank nodes, and find paths between nodes in a network. In short, a property graph model is designed to carry powerful graph analytics (i.e. path calculation, page rank, and centrality).

In this tutorial we introduce and compare these two graph data models, their underlying features, and general use cases. Instead of comparing different vendor products, this session will provide examples of both commercial and open source products. Specific feature comparisons will include: deployment options (database, in-memory, Hadoop), access methods, search, and analytical capabilities. The session will also highlight how it is possible to combine RDF and property graph platforms to support rich analytic workflows.