We present a new multidimensional model with textual dimensions based on a knowledge structure extracted from the texts, where any textual attribute in a database can be processed, and not only XML texts. This dimension allows to treat the textual data in the same way as the non-textual one in an automatic way, without user's intervention, so all the classical operations in the multidimensional model can been defined for this textual dimension. While most of the models dealing with texts that can be found in the literature are not implemented, in this proposal, the multidimensional model and the OLAP system have been implemented in a software tool, so it can be tested on real data. A case study with medical data is included in this work.

Abstract. In this work, we present a proposal of a new multidimensional model handling semantical information coming from textual data. Based of a semantical structures called AP-structures, we add new textual dimensions to our model. This new dimension aloow the user to enrich the data analisis not only using lexical information (a set or terms) but the meaning behind the textual data.