: In today’s technology driven world where most information is stored digitally, being able to efficiently retrieve and sort the available information becomes highly crucial. Supervised Learning provides solution to some of these tasks, but they are highly dependent on human interaction for categorization of texts and other attributes in the initial stages. Unsupervised Learning methods on the other hand not only provide a viable solution to some of these tasks, but they also decrease the time taken to perform various tasks by removing the need for human interaction. In this paper we shall discuss an efficient way for text mining using feature extraction, feature selection, clustering and cluster evaluation. We shall then go on to discuss an efficient way of recognizing objects using Markov’s Random Field.