Database Reference
In-Depth Information
CONCLUSION
In real life multidatabases it is not always desirable to perform complete instance integration, as there
are global users/ applications that require a global schema just to query the local databases. For them it
is important to identify the source of instances in order to make decisions. Hence in this work, we have
described an extended relational model in the context of fuzzy multidatabases with instance integration
(level-0) performed on export fuzzy relations. While integrating instances, semantic conflicts are resolved
suitably and information about the source database identity is attached with each of the resulting fuzzy
tuples. A set of such fuzzy tuples having source information attached with them are called Fuzzy Tuple
Source (FTS)-relation. A set of such FTS relations form a fuzzy multidatabase of type-2 under FTS
relational model as per our proposal. We have proposed and implemented a full set of FTS relational
algebraic operations capable of manipulating an extensive set of fuzzy relational multidatabases of type-2
that include fuzzy data values in their instances.
In this chapter we have also proposed a fuzzy query language FTS-SQL to formulate a global fuzzy
query on a fuzzy relational multidatabase of type-2 under FTS relational model.
FTS relational opera-
tions
operate on FTS relations to produce a resultant FTS relation. We have also provided architecture
for distributed fuzzy query processing with a strategy for fuzzy query decomposition and optimization.
Sub-queries obtained as a result of fuzzy query decomposition are allowed to get processed in parallel
at respective local fuzzy databases. This reduces the fuzzy query processing time effectively. We have
proved the correctness of FTS relational data model by showing that the FTS relational operations
are consistent with the fuzzy relational operations on export fuzzy relations of component local fuzzy
relational databases. Finally, we have described some algebraic properties of the FTS relational model
that may help the fuzzy query processor to transform the relational expression in order to obtain an al-
gebraically equivalent relational expression that require the least evaluation cost in terms of disk space
and communication overhead.
SCOPE FOR FUTURE WORK
The future work will explore an appropriate extension to the relational model for integrated fuzzy
databases with level-1 instance integration using information on fuzzy inclusion dependencies among
component fuzzy relational databases. A user-friendly fuzzy query interface can be designed and devel-
oped for FTS-SQL queries in the heterogeneous database environment.
REFERENCES
Agrawal, S., Keller, A. M., Wiederhold, G., & Saraswat, K. (1995). Flexible relation: An approach for
integrating data from multiple, possibly inconsistent databases”,
In: Proc, Intl. Conf. on Data Engineer-
ing
, pp 495-504.
Batini, C., Lenzerini, M., Navade, S.B. (1986). A coperative analysis of methodlogies for database
schema integration.
ACM Computing Surveys
,
18
(4), 323-364.
