Web Document Clustering Using KEA-Means Algorithm

In most traditional techniques of document clustering, the number of total clusters is not known in advance and the cluster that contains the target information or précised information associated with the cluster cannot be determined. This problem solved by Kmeans algorithm. By providing the value of no. of cluster k. However, if the value of k is modified, the precision of each result is also changes. To solve this problem, this paper proposes a new clustering algorithm known as KEA-Means algorithm which will combines the kea i.e. key phrase extraction algorithm which returns several key phrases from the source documents by using some machine learning language by creating model which will contains some rule for generating the no. of clusters of the web documents from the dataset and the k-means algorithm. This algorithm will automatically generate the number of clusters at the run time. This KEAMeans clustering algorithm provides easy and efficient way to extract test documents from massive quantities of resources.