Towards OLAP - Based Data Mining Using Multidimensional Database and Fuzzy Decision Trees
In this paper, a new approach for data mining is described, coupling fuzzy multi-dimensional databases and fuzzy data mining systems, and achieving knowledge discovery from imperfect data. An architecture based on fuzzy multidimensional databases is given. It uses these kinds of data repositories to extract relevant knowledge from large data sets from the real world. According to several works that highlighted the great interest of the OLAP framework in the knowledge discovery process, this approach enhances the existing solutions. It provides a way to deal with data from the real world, and to apply flexible operations on data sets stored as multidimensional arrays, generating more understandable fuzzy rules.