A Fuzzy Data Mining Application Over User Knowledge Involving Fuzzy Resemblance Relations

Data mining techniques have been proven as effective tools in the search of hidden or implicit knowledge in a database. Commonly, real world problems must be modeled, for example, by means of fuzzy logic, resulting in flexible database models applications and extensions of existing analysis techniques and tools. In this work, a methodology based on fuzzy data mining is applied on a particular problem involving agricultural data, such as soil information and cultivation practices, where imprecision and uncertainty factors are mainly present. We propose the application of approximate reasoning tools as linguistic terms and fuzzy resemblance relations (less restrictive than similarity relations) and we discuss the obtained results from domain experts point of view.