How ibm/digital-twin-benchmark-model works
DTBM is a collection of scripts for benchmarking digital-twin knowledge graphs and comparing them with LUBM and BSBM across graph backends. The main workflow is to choose a benchmark and backend, run the backend-specific scripts (which configure containers and execute tests), then collect benchmark results. DTBM model generation and the KITT-specific RDF import and BSBM query tools are distinct parts of that workflow. Backend script internals were not sampled, so those links are based on README statements rather than inspected calls.
Main entry points: lubm_graphdb.sh LUBM runner · graphdb GraphDB scripts · kitt KITT scripts · factory_graphdb.sh DTBM runner
Components 19
Benchmark workloads
- DTBM model
dt_model_generator.py - LUBM workload
LUBM_Benchmark.py - BSBM queries
kitt/bsbm_exp.py
Benchmark runners
- DTBM runner
graphdb/factory_graphdb.sh - LUBM runner
graphdb/lubm_graphdb.sh - BSBM runner
graphdb/bsbm_graphdb.sh - LUBM query executor
LUBM_Benchmark.py - LUBM RDF cleanup
fix_lubm.py
Graph backends
- Blazegraph scripts
blazegraph - GraphDB scripts
graphdb - Virtuoso scripts
virtuoso - Neo4j scripts
neo4j - KITT scripts
kitt
KITT tools
- KITT LUBM runner
kitt/LUBM_Benchmark.py - RDF importer
kitt/import_rdf.py - LUBM RDF importer
kitt/import_lubm_rdf.py
Other components
- Benchmark user
- Graph backend
- Benchmark results
Connections 22
- Benchmark user → DTBM runner: runs DTBM
README.md - Benchmark user → LUBM runner: runs LUBM
README.md - Benchmark user → BSBM runner: runs BSBM
README.md - DTBM runner → DTBM model: generates model
README.md - DTBM runner → GraphDB scripts: configures backend
README.md - LUBM runner → LUBM workload: executes workload
README.md - LUBM runner → GraphDB scripts: configures backend
README.md - BSBM runner → GraphDB scripts: configures backend
README.md - LUBM query executor → Benchmark results: writes timings
LUBM_Benchmark.py - LUBM query executor → Graph backend: queries endpoint
LUBM_Benchmark.py - LUBM RDF cleanup → LUBM runner: prepares RDF
- Blazegraph scripts → Graph backend: runs backend
- GraphDB scripts → Graph backend: runs backend
- Virtuoso scripts → Graph backend: runs backend
- Neo4j scripts → Graph backend: runs backend
- KITT scripts → Graph backend: runs backend
- KITT scripts → KITT LUBM runner: runs LUBM
- KITT scripts → BSBM queries: runs queries
- RDF importer → Graph backend: uploads RDF
kitt/import_rdf.py - LUBM RDF importer → Graph backend: uploads RDF
kitt/import_lubm_rdf.py - KITT LUBM runner → Graph backend: posts queries
kitt/LUBM_Benchmark.py - KITT LUBM runner → Benchmark results: writes timings
kitt/LUBM_Benchmark.py