Skip to content

Benchmark categories

In RiverBench, benchmark tasks and profiles are grouped into benchmark categories, listed below. A benchmark task is simply a description (a recipe) for conducting benchmarks. Profiles are collections of datasets, grouped by shared technical characteristics (e.g., stream type, whether RDF-star is used or not). Within each category, each task may be used with any of the profiles.

Categories

  • Flat RDF (sequences of triples/quads) – benchmarks for flat RDF collections (RDF graphs, RDF datasets) that can also be viewed as sequences of triples or quads.
    • Example task: compressing an RDF graph.
  • Grouped RDF streaming – benchmarks for streams of RDF graphs or datasets.
    • Example task: measuring the throughput of streaming RDF datasets over the network.

See also