Database Reference
In-Depth Information
Table 6. The fuzzy relation EmpDB
(Age=young Salary=reasonable)
Id
Salary
Age
…
γ
Age=
young
Salary=
reasonable
1
45000
62
0
2
38750
24
0.5
3
37500
28
0.25
set difference operations are considered. Special attention is also paid to the division operation which
may be interpreted in a different way due to many possible versions of the implication available in fuzzy
logic (Bosc, Pivert & Rocacher, 2007). Other operations typical for SQL are also redefined, including
the 'GROUP BY' clause, the 'HAVING' clause and the operators 'IN' and 'NOT IN' used along with
sub-queries. A query in SQLf language has the following syntax:
SELECT [
n|t|n,t
] set of attributes
FROM set of relations
WHERE set of fuzzy predicates
where the parameters
n
and
t
of the select block limit the number of the answers by using a quantita-
tive condition (the best
n
answers) or a qualitative condition (ones which satisfy the fuzzy predicates
according to a degree higher than
t
).
For more details and more general sources about fuzzy querying, please refer to (Galindo, 2008).
2.3.3 Keyword Search
Keyword-searchable database systems offer a simple keyword-based search interface where users need
to neither understand the underlying database schemas and structures in advance nor know complex
query languages like SQL. Instead, users are only required to submit a list of keywords, and the system
will return ranked answers based on their relevance to query keywords.
Consider the example of Figure 10, which illustrates a small subset of the DBLP database. If a user
wants to get the papers co-authored by '
Hristidis Vagelis
' and '
Papakonstantinou Yannis
', she/he should
learn the schema of the DBLP database first, and then she/he must write intricate SQL queries like this:
(
SELECT title
FROM Paper, Writes, Author
WHERE Paper.pid = Writes.pid
AND Writes.aid = Author.aid
AND Author.name = '
Hristidis Vagelis
'
) (
SELECT title
FROM Paper, Writes, Author
WHERE Paper.pid = Writes.pid
AND Writes.aid = Author.aid
