OntoRelα

Application for generating a relational database using an ontology

OntoRelα generates from an ontology and a set of configuration files, a set of scripts that create a fully functional relational database (RDB). The method consists of two processes: the ontological analysis and the ontology conversion. The analysis process verifies the structure of the ontology and builds an appropriate internal representation of the ontology. The conversion process converts the ontology constructs into a logical model as SQL scripts that creates the RDB.

Ontorelα generates from an OWL ontology and some configuration files: (1) scripts for the RDB (RDB scripts), (2) a list of warnings, (3) a mapping catalog between ontologixal construct and relational construct (OntorelCat) and (4) a normalized ontology formalized according to μOnto (MOnto). μOnto is an abstract notation inspired by the set theory and the description logic ALCQI(D) needed for the logical formulation of a relational data model (Chomicki and Saake, 1998).

References

Khnaisser C, Lavoie, L., Fraikin, B., Barton, A., Dussault, S., Burgun, A., and Ethier J.F. Using an ontology to derive a sharable and interoperable relational data model for heterogeneous healthcare data (submitted to ADBIS 2021). 2020. A summary can be found here http://griis.ca/horg-ontorela/

Khnaisser, C, Lavoie L., Burgun A., and Ethier J.F. Generating Relational Database Using Ontology Review : Issues, Challenges and Trends. International Journal of Advanced Computer Science and Applications (IJACSA) 9, nᵒ 6 (2018): 139--145. https://doi.org/10.14569/IJACSA.2018.090620.

Khnaisser, Christina. « Construction de modèles de données relationnels temporalisés guidée par les ontologies ». Université de Paris cotutelle Université de Sherbrooke, 2019. https://savoirs.usherbrooke.ca/handle/11143/16329.

Environnement

OntoRelα is developed in Java 8 with the following external libraries:

• OWLAPI 5.1 for loading and analysing the ontology.
• JGraphT to store a graph representation of the ontology and the RDB.
• Snakeyaml to read YAML configuration files.
• Jackson to write JSON files.
• StringTemplate for SQL generation. For now, only PostgreSQL are supported.
• Logback for logging.

Inputs

An ontology configuration (ontology_config.yml) to parameterize the analysis process. It consists of a list of classes and properties of interest and a list of OWL annotations properties. The μOnto verification generates a list of warnings which includes:

• the list of all classes and properties without labels or definition annotations,
• the list of all classes without a data association axiom,
• and the messages issued by the redundancy reduction step. Moreover, μOnto can be filtered using a subset of interest classes and interest properties to generate specific RDB.

An RDB configuration (database_config.yml) to parameterize the conversion process. It consists of the target RDBMS, the name of the database schema, the types of dbid and dtid attributes, the maximum length of table identifiers, data type conversion method, a list of natural languages of interest, and a list of correspondences between OWL types and SQL types according to the target RDBMS.

See content of demo for examples.

Outputs

OntoRelα generates the following outputs:

• DatabaseScripts : repository for the SQL scripts.

• • Graphs
  • InitialOntoGraph.dot : initial ontology graph.
  • OntoGraph.dot : normalized ontology graph.
  • OntoRelGraph.dot : ontoRel graph.
  • RelGraph.dot : database relational graph.

• DatabaseIdentifierAnomalies.txt : maximum length of an identifier (60 characters for PostgreSQL).

• DatabaseReport.txt : database content summary.

• ontologyname.owl_normalized.txt : normalized ontology formalized with μOnto.

• ontologyname.owl_normalized.ttl : normalized ontology formalized with Turtle.

• MOnto-summary.txt : summary of the ontology builder process.

• OntologyAnnotationDiagnostics.txt : ontology missing annotations warnings.

• OntologyAnomalies.txt : missing axioms with data properties for leaf classes.

• OntologySummary.txt : ontology content summary.

• OntoRelA-summary.txt : quantitative summary of the ontology content.

• OntoRelDiagnostics.txt : warnings for the reduction process.

• README.txt : summary of OntoRelα execution.

See the content of demo for examples.

How to use

1. First make sure you have Java installed, Java 1.8+ JDK is required.
2. Download th zip or clone this repository.
3. Run the following command in a terminal: java -jar OntoRelA-1.0.0.jar path-To-A-Configuration-Directory For example, to generate an RDB schema for the MONDIAL ontology provided in the demo repository run: java -jar OntoRelA-1.0.0.jar demo/MONDIAL/config00

Work in progress

The current version can be used with :

• OWL ontology as an input.
• PostgreSQL v9.5+ to execute de SQL files.

Source code will be available soon.

Repository content

demo : a dataset of various ontologies.
results : contain OntoRelα 1.0.0 output for ontologies demo.
report : contain technical reports.
OntoRelA-1.3.0.jar : the executable file of the application.

Responsables

For more information please contact :

Christina Khnaisser : [email protected]
Luc Lavoie : [email protected]
Jean-François Ethier : [email protected]

Copyright et licenses

Copyright 2016-{localyear}, GRIIS GRIIS (Interdisciplinary Research Group in Health Informatics)
Faculté des sciences et Faculté de médecine et sciences de la santé
Université de Sherbrooke (Québec) J1K 2R1
CANADA

The software is licensed under LILIQ-R+ version 1.1

The documentation is licensed under Creative Common CC BY SA