Title

Accommodating Instance Heterogeneities in Database Integration

Publication Type

Journal Article

Publication Date

11-2004

Abstract

A complete data integration solution can be viewed as an iterative process that consists of three phases, namely analysis, derivation and evolution. The entire process is similar to a software development process with the target application being the derivation rules for the integrated databases. In many cases, data integration requires several iterations of refining the local-to-global database mapping rules before a stable set of rules can be obtained. In particular, the mapping rules, as well as the data model and query model for the integrated databases have to cope with poor data quality in local databases, ongoing local database updates and instance heterogeneities. In this paper, we therefore propose a new object-oriented global data model, known as OORA, that can accommodate attribute and relationship instance heterogeneities in the integrated databases. The OORA model has been designed to allow database integrators and end users to query both the local and resolved instance values using the same query language throughout the derivation and evolution phases of database integration. Coupled with the OORA model, we also define a set of local-to-global database mapping rules that can detect new heterogeneities among databases and resolve instance heterogeneities if situations permit.

Discipline

Databases and Information Systems

Research Areas

Data Management and Analytics

Publication

Decision Support Systems

Volume

38

Issue

2

First Page

213

Last Page

231

ISSN

0167-9236

Identifier

10.1016/S0167-9236(03)00103-9

Publisher

Elsevier

Additional URL

http://dx.doi.org/10.1016/S0167-9236(03)00103-9