There are so many methods to cluster the software components described in data mining. Then how to know about the best method. For this there should be some evaluation methods. I want to know about best method.
My problem is not the clustering but how to evaluate the results of clusters. How we came to know that the clusters formed are quality clusters or not.
Good, you compare your final cluster result with statistical test result (T test, Chi square test, or correlation test) which is suitable for your application,