The k-means strategy is a generally utilized clustering procedure that tries to minimize the averagesquared distance between focuseson a similar group. Despite the fact that it offers no exactness ensures,its effortlessness, as well as speed, are extremely engaging. Open source products/projects focusingon the same or comparable applications are basic these days.Software maintenance is very overexerted and timeconsuming process.It utilizes lots of imperative benefits during software development process.Consequently, to design objectbased software framework, it is essential to recognize maintainability on the top priority of software development. To discover maintainable classes various machine learning techniques are applied. In this paper, K-Means clustering method is utilized at the design level to identify the maintainable classes, and the descriptive analysis has demonstrated in terms of performance metric as confusion matrix.