ProvGen: generating synthetic PROV graphs with predictable structure

TL;DR

provGen is a tool for generating large, customizable synthetic provenance graphs with predictable structures, useful for testing systems and researching graph algorithms.

Contribution

introduces provGen, a novel generator for large, structured synthetic provenance graphs with user-controlled properties and a method for evaluating their realism.

Findings

provGen can produce large graphs with specified topological features

the generated graphs effectively simulate real-world provenance patterns

the tool supports controlled testing of provenance management systems

Abstract

This paper introduces provGen, a generator aimed at producing large synthetic provenance graphs with predictable properties and of arbitrary size. Synthetic provenance graphs serve two main purposes. Firstly, they provide a variety of controlled workloads that can be used to test storage and query capabilities of provenance management systems at scale. Secondly, they provide challenging testbeds for experimenting with graph algorithms for provenance analytics, an area of increasing research interest. provGen produces PROV graphs and stores them in a graph DBMS (Neo4J). A key feature is to let users control the relationship makeup and topological features of the graph, by providing a seed provenance pattern along with a set of constraints, expressed using a custom Domain Specific Language. We also propose a simple method for evaluating the quality of the generated graphs, by measuring how realistically they simulate the structure of real-world patterns.