Abstract: Web usage mining method for learning episodic web access patterns from web usage logs which integrates knowledge on customer’s interest and behaviours. A novel method to efficiently provide better web-page recommendation through semantic-enhancement by integrating the domain and web usage knowledge of a website is introduced in this study. Here, we propose, a hybrid approach by first enriching the contextual information by k-medoids algorithm and training each cluster using simple Neural Network approach. This method improves the cluster quality in term of accuracy and CPU time when compared to traditional clustering techniques like k-means clustering and deterministic genetic clustering techniques.