A framework for RDFizing query logs and benchmarking queries and graph patterns.
LSQ2 introduces significant improvements over the prior version in every aspect: Ease-of-use, flexibility, modularity, consintency in the data model and generated IDs.
- Pretty CLI (thanks to on picocli)
- Easier yet more flexible to use: RDFization, static analysis and benchmarking now decoupled
- Named graph stream approach: Information for each query is grouped in its own named graph which allows easily selecting subsets with complete information for detailed analysis.
For detailed documentation about setup, use and concepts of the LSQ command line tool please refer to our LSQ Website.
This is a typical maven project and can is thus built with mvn clean install.
For Ubuntu/Debian users: The build process creates a .deb package that can be conviently installed after build with
./reinstall-deb.sh (requires root access).
A quick reference for the typical process is as follows:
lsq rx probe file.log
lsq rx rdfize -e http://server.from/which/the/log/is/from file.log > file.log.trig
lsq rx benchmark create -d myDatasetLabel -e http://localhost:8890/sparql -o > benchmark.conf.ttl
lsq rx benchmark prepare -c benchmark.conf.ttl -o > benchmark.run.ttl
lsq rx benchmark run -c benchmark.run.ttl *.log.trig
The -o option causes the settings to be written to the console. Omit -o to have LSQ auto-generate files.
Run example running LSQ to RDFize SPARQL logs, input and output files in the current working directory (replace $(pwd) by ${PWD} for Windows PowerShell):
docker run -it -v $(pwd):/data ghcr.io/aksw/lsq rx rdfize --endpoint=http://dbpedia.org/sparql virtuoso.dbpedia.log Build the Docker image from the source code:
docker build -t ghcr.io/aksw/lsq .The source code of this repo is published under the Apache License Version 2.0.