Database Reference
In-Depth Information
Box 1. Continued
<exp>
is an expression that led to compute the value of the com-
pound predicate. The variables <x> and <y> indicate the first and second term
respectively
Fuzzy quantifiers with the syntax:
CREATE [ABSOLUTE/RELATIVE] QUANTIFIER <name> AS <fset>
where
<name>
is a string of characters
<fset>
is the specification of a Fuzzy Set with one of the follow-
ing forms:
A trapezoidal function with four parameters
(<support
1
>, <core
1
>, <core
2
>, <support
2
>)
An arithmetic expression with the variable
<x>
that indicates the
predicate´s argument.
Checks fuzzy conditions with the syntax:
CHECK(<fuzzy condition>)
Views based in fuzzy subquery:
CREATE VIEW <name> AS <fuzzy subquery>
Fuzzy Querying and Data Manipulation
Between all different SQLf querying structures and data manipulation sentences, we focus our attention
in this chapter to provide the functionality shown in Box 2. In the rest of this chapter we will refer to
this sub language as SQLf-DML.
DESIGN OF A FUZZY QUERY PROCESSOR
We propose a fuzzy query processor as the extension of an existing querying engine. For so doing, first
we must to extend the RDBMS' catalog in order to record the fuzzy terms (predicates, comparators,
modifiers and quantifiers). Thereafter, we extend the parser module in order to accept SQLf-DML and
SQLf-DDL syntax. Then a general processing algorithm involving different modules may be applied
according the kind of recognised sentence.
Statement Processing
The general algorithm for term definition is:
1. The parser accepts the creation of a
fuzzy term
(i.e. create fuzzy predicate…, create fuzzy modi-
fier…, create fuzzy quantifier …)
2. If the
fuzzy term
does not exists then record it with parameters on the fuzzy catalog, else report to
the user that the
fuzzy term is
already on the fuzzy catalog.
