Database Reference
In-Depth Information
Figure 1. Fuzzy Relational Algebra Tree
This derived Fuzzy Query Tree with boolean conditions can be processed by classical query opti-
mizer, then we propose to extend access mechanisms (sequential scan, index scan) and physical algebra
operators (projection, selection, join) such that the query evaluator might apply fuzzy sequential scan,
fuzzy index scan, fuzzy projection, fuzzy selection or fuzzy join.
The fuzzy operators can compute the degree of membership and send to the next fuzzy algebra opera-
tor the membership degree, additionally if the query optimizer push the projection below, this operator
must to project the linguistic variable (age and grade in this case) as is shown in Figure 3.
We consider for fuzzy mechanisms access and algebra operators that they may have a classical rela-
tion as input but a fuzzy relation (classical relation with a membership degree) as output. In case of join
with two fuzzy relations as input the output fuzzy relation will have the minimum between the member-
ship degrees.
Figure 2. Derived Fuzzy Query Tree
