property graph is commonly adopted by graph database community. The two models are inherently different . (1) reifying binary relation between two entities into a new node in graph will add one more join in query (2) in practice, a binary relation decorated with context like valid time interval, source, confidence, explanation, and etc, enable users to search on a big connected graph with certain filters (instead of dropping out arcs). Therefore, although RDF graph somehow offers a simpler data model, property graph seems more optimized for applications.
1.While mapping from RDF to property graph is straightforward, the reverse directly need some work. How to represent the attributes attached to each individual property graph's arc? Das et al. [1] investigated three mapping models.
property graph is commonly adopted by graph database community. The two models are inherently different . (1) reifying binary relation between two entities into a new node in graph will add one more join in query (2) in practice, a binary relation decorated with context like valid time interval, source, confidence, explanation, and etc, enable users to search on a big connected graph with certain filters (instead of dropping out arcs). Therefore, although RDF graph somehow offers a simpler data model, property graph seems more optimized for applications.
1.While mapping from RDF to property graph is straightforward, the reverse directly need some work. How to represent the attributes attached to each individual property graph's arc? Das et al. [1] investigated three mapping models.
[1] A Tale of Two Graphs: Property Graphs as RDF in Oracle