In this paper, we develop techniques to produce interoperable queries with
object and relational databases. A user poses a local query in a local quer
y language, against a local object or relational schema. We transparently p
roduce appropriate queries with respect to a remote target object or relati
onal schema, corresponding to some remote database which contains data rele
vant to the user's query. Mapping knowledge to resolve representational het
erogeneities in local and remote schemas is expressed in a canonical repres
entation, CRmapping, and is independent of the particular data model. A can
onical representation CRquery is also used to resolve heterogeneities of qu
ery languages. A set of heterogeneous transformation algorithms define the
appropriate transformations from the local queries to the remote queries. T
he use of canonical representations (CR) allows us to represent queries ind
ependent of the particular query language, and to resolve representational
conflicts in a uniform manner, independent of models and query languages.