Interpretations of Discovered Knowledge in Multidimensional Databases

It is a big challenge to guarantee the quality of discovered knowledge in multidimensional databases because of the huge amount of patterns and noises. The essential issue is to provide efficient methods for interpreting meaningful discovered knowledge in databases. This research presents a new technique called granule mining to improve the performance of data mining. Rather than using patterns, it uses granules in different tiers to generalize knowledge in databases. It also provides a mechanism to formally discuss meaningless discovered rules based on relationships between granules in different tiers.