Abstract:To overcome the Weka’s weakness in clustering algorithms, the clustering algorithm was developed based on the Weka open source environment. An introduction is made on the basic principle and construction steps of the Chameleon algorithm which was described concretely and embedded into Weka platform whose inner class and visualization function were exploited adequately. The newly built Weka system was tested by comparing the embedded and the already existed K-means algorithm. The result shows that Chameleon algorithm acquires better clustering effect.