A distributed knowledge extraction data mining algorithm

Abstract

We have developed a distributed data mining algorithm based on the progressive knowledge extraction principle. The knowledge factors, the data attributes that are significant statistically or based on a predefined mining function, are extracted progressively from the distributed data sets. The critical data attributes and sample data set are selected iteratively from distributed data sources. The experiments showed that the algorithm is valid and has the potentials for the large distributed data mining practices. © Springer-Verlag 2004.