A research on extraction method of distributed heterogeneous dataset in multi-support association rule mining

This paper proposes a method of EMD (extraction method of distributed heterogeneous dataset in multi-support association rule mining) which can be applied into filtering, abstraction, analysis and transformation of data feature record set in multi-support association rule mining. In order to ensure the efficient implementation of multi-support association rule mining, the format of data feature record set should be uniform, especially in the mining of distributed heterogeneous and mass data. Firstly, EMD utilizes XML to define extraction semantic, then extracts and filters data from corresponding data sources according to the semantic. Finally after analyzing the data and transforming it into a standard format, the data is saved in XML for multi-support association rule mining next step.