E-Learning is suitable when the learners are grouped and facilitated to learn according to their learning parameters and in their own pace. In this line the partition clustering method algorithms are identified to be appropriate and the researcher identified that the K-Medoid clustering better suits the grouping of e-Learners. Identifying the most preferred learning activities of a learner will ensure quicker learning capacities. Out of the existing dimensionality reduction methods, Principal Component Analysis (PCA) best reduces the dimensions, while preserving the integrity of the data set and hence PCA is applied to reduce the learner activities. Apart from that there is a necessity to fix the value of K (cluster size) when the K-Mediod clustering has to be extended for grouping of E-Learners. Hence, the researcher has also identified and proven that the Silhouette method best suits to fix the optimal value for K. The various stages of this research depicted above can be synthesized as a framework for implementing the steps taken throughout the research. In this research article a framework is proposed to grouping of e-Learners by using preferred e-Learning activities through benchmark algorithms.