establishments and campgrounds. In the group there
are youths (25-24) coming from Catalonia and the
Basque Country and it is the first time that visit the
Delta. The group BeachTourism (39,3%) was
characterised by their interest in beach, relaxation,
and walking. It is family tourism, staying in rental
apartments or second homes. They come from Spain
or overseas. The group contains middle-class people,
35-64 years old, who do long and frequent visits.
4 STUDY OF THE VISITORS
WITH CLUSTERING
In this section, a hierarchical clustering based on the
Ward’s criterion (Ward, 1963) is done on the same
dataset of visitors to the Ebre Delta. We have taken
the same subset of variables (4 that define the tourist
profile and the 6 that model the trip profile).
Table 1: Typology of visitors to the Ebre Delta Natural
Park with categorical features.
Class # Description
864 1 Single outlier visitor
C963 20 Long stage, between 35-early 40s years,
52% stays at second home, 75% are
Catalan people, it is not clear the first
reason to come (some of them come for
walking).
C966 72 Long stage, between 35-early 40s, 68% is
at home, half Spanish, half Catalan, have a
second residence near the park
C965 37 Long stage, higher fidelity, around 46
years, 65% home, 78% Catalan, their main
interest is gastronomy
C921 4 Long stage, more fidelity, between 35-
early 40s, 50% goes to the hotel, an
important part makes reservation, part of
the foreigners concentrated in this group
25% of the class are foreigners, they come
for recommendation of other people, 50%
Catalan, main interests: relaxation or
landscape
C936 16 Shorter stage, between 35-early 40s,
almost 60% home, 80% Catalan, main
interests: nature or business
C918 8 Youngs, under 30s, 50% stay in camping,
75% makes reservation, 50% Spanish,
education tends to be first reason
C964 817 Shorter stage, occasional visit, between 35-
early 40ss, mainly hotel, 63% Catalan,
main reasons: nature, landscape and
sightseeing
In the same way of the previous study, as most of
the variables were categorical, only equal or
different values are distinguished, leading to a poor
estimation of the similarity between responses.
Table 2: Typology of visitors using semantics.
Class # Description
C947 110 The 81% comes for nature, but also for
relax (35%), they use mainly hotels and
rural establishments (79%), they have a
reservation (95%)
C966 194 They come for relax (36%), visit the
family (14.4%), but the second reason is
mainly nature (35%), they have no hotel,
they stay at home or at a family house
(68,5%), and they have no reservation
(99%), this is a group of young people
leaving in the area, which repeat the visits
more than others.
C968 203 Short stage, around 2 days, they clearly
come for nature reasons (91.6%) and
second for relax and wildlife (43.6%), they
are in hotels or apartments (44.6%)
although they have not reservation, mainly
Spanish
C955 88 The first reason for coming is
heterogeneous (nature, relaxation, beach,
landscapes), the second is nature, they stay
in a camping (90%), the half have a
reservation, mainly Catalan and Spanish
but also concentrates a big proportion of
foreigners
C944 124 Relax and wildlife (46%) are the first
reasons for coming and second is nature
(40%), they stay at hotel or cottages
(72%), and have reservation (88%). This is
a group of slightly older people
programming the stay in hotel or
apartment, looking for relax or beach
C964 88 Wildlife and the landscape are the first
reasons for coming (67%), but also for
culture (19.5%) and the second reason is
nature, they are mainly in hotel (54%).
They are mainly Catalan or Spanish.
C957 84 Stay longer, slightly older that the rest,
nature (38%) and beach (16%) are the
main interest and second main interest is
wildlife, most of them are foreigners with
a second home, or that stay in an
apartment.
C961 84 They all come for beach, their secondary
interests are equally relaxation and nature,
they live near the park and their visit is
improvised, the stage is longer.
The clustering generates a big and heterogeneous
class (with 83.8% of the tourists) which seems to
share all type of visitors and other 7 small classes.
Therefore, although the interpretation of the small
classes is possible (see Table 1), from the point of
view of the manager, this partition is useless because
ICAART 2011 - 3rd International Conference on Agents and Artificial Intelligence
286