The EvoGen benchmark suite for evolving RDF data

Abstract

Artificial and synthetic data are widely used for benchmarking and evaluating database, storage and query engines. This is usually performed in static contexts with no evolution in the data. In the context of evolution management, the community lacks systems and tools for benchmarking versioning and change detection approaches. In this paper, we address the generation of synthetic, evolving data represented in the RDF model, and we discuss requirements and parameters that drive this process. Furthermore, we discuss query workloads in the context of evolution. To this end, we present EvoGen, a generator for evolving RDF data, that offers functionality for instance and schema-based evolution, fine-grained change representation between versions as well as custom workload generation.