A Subtractive Relational Fuzzy C-Medoids Clustering Approach To Cluster Web User Sessions from Web Server Logs

In this paper, a subtractive relational fuzzy c-medoids clustering approach is discussed to identify web user session clusters from weblogs, based on their browsing behavior. In this approach, the internal arrangement of data along with the density of pairwise dissimilarity values is favored over arbitrary starting estimations of medoids as done in the conventional relational fuzzy c-medoids algorithm. It is assumed that straightforward binary session dissimilarity measure proposed and utilized as a part of prior detailed work is not especially logical, and instinctive to speak to session dissimilarities instigated from web client's habits, interest, and expectations. The idea of augmented sessions is utilized to infer page relevance based web client’s session dissimilarity matrix. The discussed approach is applied on an augmented session dissimilarity matrix obtained from an openly accessible NASA web server log data. The produced clusters are assessed by utilizing diverse fuzzy validity measures, and results are contrasted with conventional fuzzy c-medoids clustering. Experimental results demonstrated that quality of fuzzy clusters produced by utilizing proposed subtractive relational fuzzy c-medoids clustering is superior as compared with the conventional relational fuzzy c-medoids approach.