Oscar Edgardo Romero Arredondo, , Jesus A. Gonzalez, Ivan Olmos and Rosalva Loreto, Benemerita Universidad Autonoma de Puebla
There are many interesting applications of knowledge discovery applied to real world structural domains. In this paper, we present the application of the graph-based data mining system Subdue to the "Multifunctional Spaces of the Properties in Puebla of the Angels in the XVI, XVII, and XVIII Centuries. Our experimental results show how important is the knowledge representation to successfully find relevant patterns at different levels of abstraction. We worked very close to the domain expert who used the patterns found to support some of her theories and also proposed new hypotheses to guide her research. This is a successful project where data mining techniques were used for decision support in a history domain.