A Hybrid Clustering Based Collaborative Filtering (CF) Approach

The main innovation in modern human society is the presence and influence of Information Technologies. Everybody uses different technological devices and computers for either personal, academic, professional, scientific, clinical and other needs. These, extremely numerous, groups of Information Technology's users generate very large amounts of data. A technique in recommender systems is called as Collaborative filtering. It predicts the user rating of an item by association of similar users who have similar interests.

An approach in collaborative querying is known as query clustering, which means to group similar queries automatically without using predetermined class descriptions. In this paper, we proposed an approach which works in two stages. In the first stage, the services are divided into small number of clusters for processing and later hybrid collaborative filtering algorithm is used on one of the clusters. We have compared the results by calculating the different similarity values such as functional similarity, description similarity and characteristic similarity by using the clustering method. The experimentation is done with MovieLens dataset which is available for research purpose provided by the GroupLens. The comparative analysis and comprehensive study shows that clustering based collaborative filtering algorithm puts forward for better performance among the other comparative algorithms.