Semantic Clustering Using Multiple Ontologies

Data mining tools able to semantically interpret textual or linguistic data are acquiring a growing importance. Moreover, the development of large ontologies for general and specific domains provides new tools to include background knowledge into data mining techniques such as clustering. In this paper we present an unsupervised clustering method that exploits the semantics of categorical data by means of ontologies, and it is also able to manage numerical data. Our method is able to use different ontologies in order to assess the meaning of the values during the clustering process, leading to a set of clusters with a clearer semantic interpretation in a particular domain. The influence of using one or several ontologies is analyzed by using real data collected from visitors to the Ebre Delta Natural Park, which is a protected natural reserve in Catalonia (Spain).