Randomization is an economical and efficient approach for privacy preserving data mining (PPDM). In order to guarantee the performance of data mining and the protection of individ...
Frequent pattern mining on data streams is of interest recently. However, it is not easy for users to determine a proper frequency threshold. It is more reasonable to ask users to ...
Our research aims at discovering useful knowledge from the large collection of satellite images of typhoons using data mining approaches. We first introduce the creation of the typ...
Spatio-temporal patterns extracted from historical trajectories of moving objects reveal important knowledge about movement behavior for high quality LBS services. Existing approac...
Classification, which involves finding rules that partition a given da.ta set into disjoint groups, is one class of data mining problems. Approaches proposed so far for mining cla...