Functional query optimization over object-oriented views for data integration (1999) [20 citations — 9 self]
Popular Tags
No tags have been applied to this document.
BibTeX | Add To MetaCart
@ARTICLE{Josifovski99functionalquery,author = {Vanja Josifovski and Tore Risch Fvanja and M. D. Gray},
title = {Functional query optimization over object-oriented views for data integration},
journal = {Journal of Intelligent Information Systems (JIIS},
year = {1999},
volume = {12},
pages = {12--2}
}
Years of Citing Articles
Abstract:
Abstract. AMOS is a mediator system that supports passive (non-intrusive)integration of data from heterogeneous and autonomous data sources. It is based on a functional data model and a declarative functional query language AMOSQL. Foreign data sources, e.g. relational databases, text les, or other types of data sources can be wrapped with AMOS mediators, making them accessible through AMOSQL. AMOS mediators can communicate among each other through the multi-database constructs of AMOSQL that allow de nition of functional queries and OO views accessing other AMOS servers. The integrated views can contain both functions and types derived from the data sources. Furthermore, local data associated with these view de nitions may be stored in the mediator database. This paper describes AMOS ' multi-database query facilities and their optimization techniques. Calculus-based function transformations are used to generate minimal query expressions before the query decomposition and cost-based algebraic optimization steps take place. Object identi er (OID) generation is used for correctly representing derived objects in the mediators. A selective OID generation mechanism avoids overhead by generating in the mediator OIDs only for those derived objects that are either needed during the processing of a query or have associated local data in the mediator database. The validity of the derived objects that are assigned OIDs and the completeness of queries to the views are guaranteed by system generated predicates added to the queries.
