Submissions:Pattern for re-engineering a term-based thesaurus, which follows the relationbased data model, into an ontology schema

From Odp

Jump to: navigation, search


This pattern has been certified.

Related submission, with evaluation history, can be found here

If you are a member of quality committee please visit the

evaluation section

If you are author of this proposal or you want to contribute to this pattern's review, you can:

In general, it could be useful to visit the evaluation section to have informations about the evaluation process of this proposal

Current revision ID: 11035

General information

Name Pattern for re-engineering a term-based thesaurus, which follows the relationbased data model, into an ontology schema
Problem Re-engineering a term-based thesaurus, which follows the relation-based model, to design an ontology schema.


Non-Ontological Resource

Description A non-ontological resource holds a term-based thesaurus which follows the relation-based model.

A thesaurus represents the knowledge of a domain with a collection of terms and a limited set of relations between them. The relation-based data model is a normalized structure, in which relationship types are not defined as fields in a record, but they are simply data values in a relationship record, thus new relationship types can be introduced with ease.

Graphical Representation

Diagram

Image:Relationbasedtermbased.png


Ontology

Description The generated ontology will be based on the lightweight ontology architectural pattern (AP-LW-01). Each thesaurus term is mapped to a class. For the disambiguation of the semantics of the BT, NT and RT relations among thesaurus terms the pattern relies on an external resource. In the case the external resource does not provide any relation between two terms, the pattern takes advantage of the use of logical patterns for asserting the relation partOf, subClassOf or relatedClass. For the UF/USE relations we use the logical pattern proposed by Corcho et al. suggested as best practice in the context of this antipattern: the tendency to declare two classes equivalent when in fact their labels simply express synonym.
Graphical Representation

Diagram

Image:RelationbasedtermbasedOntology.png


Process

Description 1. Identify the records which contain thesaurus terms without a broader term, within the term-term relationship entity.

2. For each one of the above identified thesaurus terms ti:

2.1. Obtain the thesaurus term within the term entity.

2.2. Create the corresponding ontology class, Ci class, if it is not created yet.

2.3. Identify the thesaurus term, tj, which are narrower terms of ti, within the term-term relationship entity.

2.4. For each one of the above identified thesaurus terms tj :

2.4.1. Obtain the thesaurus term within the term entity.

2.4.2. Create the corresponding ontology class, Cj class, if it is not created yet.

2.4.3. Using the external resource identify the semantics of the relation between Cj and Ci, and set up the relation identified.

2.4.4. Repeat from step 2.2 for cj as a new ci

2.5. Identify the thesaurus term, tr, which are related terms of ti, within the term-term relationship entity.

2.6. For each one of the above identified thesaurus term tr:

2.6.1. Obtain the thesaurus term within the term entity.

2.6.2. Create the corresponding ontology class, Cr class, if it is not created yet.

2.6.3. Using the external resource identify the semantics of the relation between Cr and Ci, and set up the relation identified.

2.6.4. Repeat from step 2.4 for cr as a new ci

2.7. Identify the thesaurus term, tq, which are equivalent terms of ti, within the term-term relationship entity.

2.8. For each one of the above identified thesaurus term tq:

2.8.1. Use the logical pattern proposed by Corcho et al.

2.8.2. Repeat from step 2.6 for cq as a new ci

Graphical Representation

Diagram

Image:Wftbthrello.png


Scenario example

Description Suppose that someone wants to build an ontology schema based on earlier version of the AGROVOC Thesaurus, which is a term-based thesaurus and it follows the relation-based model.


Example of a Non-Ontological Resource

Description The AGROVOC Thesaurus is an structured and controlled vocabulary designed to cover the terminology of all subject fields in agriculture, forestry, fisheries, food and related domains.
Graphical Representation

Diagram

Image:RelationbasedtermbasedAgrovoc2.png

Web Reference http://www.fao.org/agrovoc/


Ontology example

Description The generated ontology will be based on the lightweight ontology architectural pattern (AP-LW-01). Each thesaurus term is mapped to a class. For the disambiguation of the semantics of the BT, NT and RT relations among thesaurus terms the pattern relies on an external resource. In the case the external resource does not provide any relation between two terms, the pattern takes advantage of the use of logical patterns for asserting the relation partOf, subClassOf or relatedClass. For the UF/USE relations we use the logical pattern proposed by Corcho et al. suggested as best practice in the context of this antipattern: the tendency to declare two classes equivalent when in fact their labels simply express synonym.
Graphical Representation

Diagram

Image:relationbasedtermbasedSubClassOfExample.png

Web Reference http://droz.dia.fi.upm.es/ontologies


Process example

Description 1. Create the Poaceae class.

1.1. Create the Oryza class.

1.2. Using the external resource identify the semantics of the relation between Oryza and Poaceae, and set up the relation identified.

1.2.1. Create the Rice class.

1.2.2. Using the external resource identify the semantics of the relation between Rice and Oryza, and set up the relation identified.

2. Create the Cereals class.

2.1. Using the external resource identify the semantics of the relation between Rice and Cereals, and set up the relation identified.

Graphical Representation

Diagram

Image:Wftbthrelloexample.png


About

SubmittedBy BorisVillazón-Terrazas
Author BorisVillazón-Terrazas
Also known as
Known uses
Related to
Other References

Additional information

Scenarios

Scenarios about Pattern for re-engineering a term-based thesaurus, which follows the relationbased data model, into an ontology schema

No scenario is added to this Content OP.

Reviews

Reviews about Pattern for re-engineering a term-based thesaurus, which follows the relationbased data model, into an ontology schema

There is no review about this proposal. This revision (revision ID 11035) takes in account the reviews: none

Other info at evaluation tab


Modeling issues

Modeling issues about Pattern for re-engineering a term-based thesaurus, which follows the relationbased data model, into an ontology schema

There is no Modeling issue related to this proposal.


References

Add a reference


Personal tools
Quality Committee
Content OP publishers