PG-Keys: Keys for Property Graphs

Abstract

We report on a community effort between industry and academia to shape the future of property graph constraints. The standardization for a property graph query language is currently underway through the ISO Graph Query Language (GQL) project. Our position is that this project should pay close attention to schemas and constraints, and should focus next on key constraints.The main purposes of keys are enforcing data integrity and allowing the referencing and identifying of objects. Motivated by use cases from our industry partners, we argue that key constraints should be able to have different modes, which are combinations of basic restriction that require the key to be exclusive, mandatory, and singleton. Moreover, keys should be applicable to nodes, edges, and properties since these all can represent valid real-life entities. Our result is PG-Keys, a flexible and powerful framework for defining key constraints, which fulfills the above goals.PG-Keys is a design by the Linked Data Benchmark Council's Property Graph Schema Working Group, consisting of members from industry, academia, and ISO GQL standards group, intending to bring the best of all worlds to property graph practitioners. PG-Keys aims to guide the evolution of the standardization efforts towards making systems more useful, powerful, and expressive.

Discovery of Keys for Graphs [Extended Version]

“…Graph matching keys, referred to as GMKs, are extension of graph keys using similarity predicates on values, and supporting approximation entity matching [12]. [7] proposes a modular and flexible model to formalise keys for property graph. Their keys are defined to be used for a property graph query language that is currently underway through the ISO Graph Query Language (GQL) project.…”

Threshold Queries in Theory and in the Wild

“…, 𝑛}. 1 The set of answers to 𝑞 on 𝐷 is 𝑞(𝐷) = 𝜋 FVar(𝑞) (𝜂) 𝜂 is a match for 𝑞 in 𝐷 . Notice that we define answers as functions instead of database tuples, as this simplifies the presentation and as reasoning about their underlying domains is useful in further sections, when dealing with query decompositions.…”

“…Then we sample (𝑎, 𝑚), (𝑎, 𝑛) ∈ 𝑅 {𝑥,𝑦 } × 𝑅 {𝑥,𝑧 } such that 𝑚 • 𝑛 ≥ 3 with weights given by the product of the multiplicities of (𝑎, 𝑚) in 𝑅 ′ {𝑥,𝑦 } and (𝑎, 𝑛) in 𝑅 ′ {𝑥,𝑧 } . In our case, (𝑎 1 , 1), (𝑎 1 , 3) is chosen with probability 1 2 , and both (𝑎 1 , 2), (𝑎 1 , 3) and (𝑎 1 , 2), (𝑎 1 , 2) with probability 1 4 . Finally, we sample (𝑎, 𝑏, 𝑚) and (𝑎, 𝑐, 𝑛) uniformly among relevant triples in 𝑅 {𝑥,𝑦 } and 𝑅 {𝑥,𝑧 } , respectively, and we return (𝑎, 𝑏, 𝑐).…”

“…Finally, we sample (𝑎, 𝑏, 𝑚) and (𝑎, 𝑐, 𝑛) uniformly among relevant triples in 𝑅 {𝑥,𝑦 } and 𝑅 {𝑥,𝑧 } , respectively, and we return (𝑎, 𝑏, 𝑐). In our case, for (𝑎 1 , 1), (𝑎 1 , 3) we choose either (𝑎 1 , 𝑏 1 , 1), (𝑎 1 , 𝑐 1 , 3) or (𝑎 1 , 𝑏 2 , 1), (𝑎 1 , 𝑐 1 , 3) with probability 1 2 , and for (𝑎 1 , 2), (𝑎 1 , 3) and (𝑎 1 , 2), (𝑎 1 , 2) there is only one choice; overall, each answer is returned with probability 1 4 .…”

“…The query is employed to identify pairs of drivers and clients who perform a number of rides together that is large enough to raise suspicions of collusion. The query, which is written in Cypher for illustrative purposes, can be seen as a graph cardinality constraint, and thus as a generalization of key constraints for property graphs [1].…”