Databases Reference
In-Depth Information
Chapter 8
Mapping-Based Merging of Schemas
Rachel Pottinger
Abstract
Merging schemas or other structured data occur in many different data
models and applications, including merging ontologies, view integration, data inte-
gration, and computer supported collaborative work. This paper describes some of
the key works in merging schemas and discusses some of the commonalities and
differences.
1
Introduction
Schemas, ontologies and other related structures commonly need to be merged
in a number of different applications. This happens for a number of reasons. For
example:
View integration:
Different users have their own aspects of a common application
that they are interested in. For example, in creating a database for a university,
the registrar has a different view from a professor, and both have different views
from a student. In view integration, each user group creates its own “view” of
what should be in the schema and then these different views are combined to
create one global schema in which the data is stored.
Data integration:
Users may want to query over multiple databases. For exam-
ple, a BioMedical researcher may want to query both HUGO and OMIM for
information on genes, and then use the gene information to query SwissProt for
which proteins those genes encode. Because the researcher does not want to learn
each of the schemas, and yet creating a warehouse of the entire set of databases is
infeasible because of size and access restrictions, the user would like to just query
one schema once and have the system figure out how to translate the queries over
the sources. Such a system is called a data integration system.
R. Pottinger
University of British Columbia, 201-2366 Main Mall, Vancouver, BC, Canada V6T 1Z4
e-mail:
rap@cs.ubc.ca
