IMDb-MO

Here we provide the detailed information for the IMDb-Movie Ontology experiments (in Section 3.2).

Inputs

The Movie Ontology (MO) is developed, adapted and maintained by the Department of Informatics at the University of Zurich:

• movieontology.owl (the ontology used for these experiments)

• dbpedia_3.7.owl (required as an "import"  because MO uses some of the dbpedia vocabulary)

The mapping (in Ontop's mapping syntax; to transform it into R2RML use Ontop's command-line tools)

• ontop-mappings.obda (text file)

The data for the experiments was generated using IMDb's dumps and Python scripts located at:

• http://www.imdb.com/interfaces (data)

• http://imdbpy.sourceforge.net/docs/README.sqldb.txt (script information)

To save time, you can use the data we prepared for the experiments:

• imdb-mysql-dump.sql.gz (SQL dump)

• imdb-triples.zip (data in the form of N3 RDF triples to be used for Stardog, OWLIM or any other triple store)

The list of SPARQL queries used in the experiments (the same for all systems):

• imdb-queries.txt (text file)

Summary

Query execution time (in seconds) for each query in the IMDb-MO scenario. 

These are the averages of 5 warm executions.

https://docs.google.com/spreadsheet/ccc?key=0Al3T8WsgEbGtdHZMQ1R2ZkxFYy02dVg0dkxreDVwa0E&usp=sharing

Detailed results

Detailed results for each system, including the loading time (SQL or triple) and the forward chaining time (for OWLIM):

• Ontop-DB2 settings and results

• Ontop-MySQL settings and results

• OWLIM settings and results

• Stardog settings and results

SQL queries produced by Ontop for each of the SPARQL queries:

• imdb-queries-quest-sql.txt