Meta data
- BibTeX key
- ghawi2007database
- entry type
- inproceedings
- booktitle
- Third International Workshop on Database Interoperability (InterDB 2007)
- year
- 2007
- posted-at
- 2009-11-05 12:53:43
- priority
- 0
- citeulike-article-id
- 6074141
- comment
- Good background overview in section 2, and related work in section 6. There are 2 types of database-to-ontology mappings: 1. creating ontology from db, i.e. simply mapping the tables to classes, the rows to subject, the columns to predicates and the cells to objects. 2. mapping a db to an existing ontology; requires more complex mapping. A system is proposed that combines different heterogenic datasources and provides a query Webservice to perform semantic queries as if it were one source. To achieve this, is proposes to use a central domain-model ontology. A query based on the central ontology is analyzed and distributed over the various datasources, as the central query-Webservice knows which data is available where. The different datasources have two mappings. One is the mapping of their local data representation (e.g. relational db) to a local ontology. The second is the mapping from the local ontology to the central ontology. The only part detailed enough to be useful is the automatic mapping generation (generating from db) method used by DB2OWL. They say DB2OWL uses the R2O language to define the resulting mapping description. The mapping rules is explained in detail, but is actually stolen from Stojanovic et al., 2002, with the exception of a few detailes that weren't even possible for Stojanovic because it was written in the pre-OWL era. Although this work is referred as 'related work', shame on the authors for giving so little credit to Stojanovic! The good news is that they do a better job at explaining the method. The document does mention a similarity estimation process to estimate how good two ontologies match, but does not elaborate on this process nor the mapping of the local ontology to the central ontology - while this is the hard part! I should find a source that explains the rest, if that has been actually worked out. (Stojanovic et al, 2002): Ljiljana Stojanovic, Nenad Stojanovic, Raphael Volz. Migrating data-intensive web sites into the Semantic Web; In Proceedings of the 2002 ACM Symposium on Applied Computing (2002), pp. 1100-1107.
Tags
Users
Comments and Reviewsshow / hide
Cite this publication
More citation styles
- please select -%0 Conference Paper
%1 ghawi2007database
%A Ghawi, R.
%A Cullot, N.
%B Third International Workshop on Database Interoperability (InterDB 2007)
%D 2007
%K RDB2RDF db2owl
%T Database-to-Ontology Mapping Generation for Semantic Interoperability
@inproceedings{ghawi2007database,
added-at = {2011-02-07T11:30:23.000+0100},
author = {Ghawi, R. and Cullot, N.},
biburl = {https://www.bibsonomy.org/bibtex/2c8e40bf51b5433b5e6377fa1f843ca5a/sebastian},
booktitle = {Third International Workshop on Database Interoperability (InterDB 2007)},
citeulike-article-id = {6074141},
comment = {Good background overview in section 2, and related work in section 6.
There are 2 types of database-to-ontology mappings:
1. creating ontology from db, i.e. simply mapping the tables to classes, the rows to subject, the columns to predicates and the cells to objects.
2. mapping a db to an existing ontology; requires more complex mapping.
A system is proposed that combines different heterogenic datasources and provides a query Webservice to perform semantic queries as if it were one source. To achieve this, is proposes to use a central domain-model ontology. A query based on the central ontology is analyzed and distributed over the various datasources, as the central query-Webservice knows which data is available where. The different datasources have two mappings. One is the mapping of their local data representation (e.g. relational db) to a local ontology. The second is the mapping from the local ontology to the central ontology.
The only part detailed enough to be useful is the automatic mapping generation (generating from db) method used by DB2OWL. They say DB2OWL uses the R2O language to define the resulting mapping description. The mapping rules is explained in detail, but is actually stolen from Stojanovic et al., 2002, with the exception of a few detailes that weren't even possible for Stojanovic because it was written in the pre-OWL era. Although this work is referred as 'related work', shame on the authors for giving so little credit to Stojanovic! The good news is that they do a better job at explaining the method.
The document does mention a similarity estimation process to estimate how good two ontologies match, but does not elaborate on this process nor the mapping of the local ontology to the central ontology - while this is the hard part! I should find a source that explains the rest, if that has been actually worked out.
(Stojanovic et al, 2002): Ljiljana Stojanovic, Nenad Stojanovic, Raphael Volz. Migrating data-intensive web sites into the Semantic Web; In Proceedings of the 2002 ACM Symposium on Applied Computing (2002), pp. 1100-1107.},
interhash = {ffdad93ef20affc5cf206eb9cc535aff},
intrahash = {c8e40bf51b5433b5e6377fa1f843ca5a},
keywords = {RDB2RDF db2owl},
posted-at = {2009-11-05 12:53:43},
priority = {0},
timestamp = {2011-02-07T11:30:44.000+0100},
title = {{Database-to-Ontology Mapping Generation for Semantic Interoperability}},
year = 2007
}