The Research and Selection of Ideal Cloud Services using Clustering Techniques

Abstract

The appropriate selection of cloud services is a new need that resulted from the increased use of Cloud Computing. Users want to find the cloud services that best meet their requirements from a large number of available cloud services. This requires a special treatment. Generally, it is done by using various techniques, such as Similarity, Data Mining, Recommender Systems, Multi-Criteria Decision Analysis (MCDA) methods, etc. The main problem of these cited methods, is firstly the large number of returned services and secondly the absence of direct communication between the user and the system. Most of methods are based on previous system knowledge, or they rely on other users feedback. In order to avoid these weaknesses we thought to add a form of communication between the user and the system. For this end, in our contribution, the user is requested to enter the values of his ideal service and then based on the techniques of clustering, the services which are the most closest to the ideal service are selected and returned to the final user.