Database Reference
In-Depth Information
Table 10. Definition of INTERVAL_VALUE ENTITY and its instantiation
INTERVAL_VALUE ENTITY
ENTITY INTERVAL_VALUE
min_value: REAL;
max_value: REAL;
WHERE min_value < max_value;
END_ENTITY;
Table 11. Definition of INTERVAL PREFERENCE and its instantiation
INTERVAL_PREFERENCE
ENTITY INTERVAL_PREFERENCE
SUBTYPE OF (INTERPRETED_PREFERENCE);
interpreted_by: LIST[1:?] INTERVAL_VALUE;
pref_attributes: LIST[1:?] OF PREFERENCE_URI;
END ENTITY;
Figure 5. Ontology preference relationship
Context Based Preferences
#26 instance states that a probability of 0.1
is associated to the cheap preference. PROB_
VALUE is a redefinition of the REAL data type
of EXPRESS.
The definition of preferences may depend on
the context in which they can be interpreted. For
example, someone may interpret differently the
cheap preference attached to the price of a hotel
depending on whether the hotel is localised in
Paris, London, or Poitiers. In this case, the inter-
pretation of a preference depends on the value of
another property ( localisation in our example).
To handle these kinds of preferences, we have
introduced the notion of context preferences de-
fined in EXPRESS as indicated in Table 14. The
PROPERTY_VALUE of price correspond to the
instance property #p2 presented in Figure 7.
Uninterpreted Preferences
Uninterpreted preferences correspond to a set
of ontology's properties or classes instances
which are considered as preferred. There is no
any rationale for choosing these instances. For
example, HotelIBIS, SwimmingPool, pizza, and
coffee can be the different preferences that one
can have in the tourism domain. The EXPRESS
code defining such preferences and the instance
#47 corresponding to HotelIBIS, SwimmingPool
instances are given in Table 13.
 
Search WWH ::




Custom Search