A Classification Framework Towards Application of Data Mining in Collaborative Filtering

Research Methodology

In this chapter, various collaborative filtering techniques are studied for improving the quality of the recommendation process. The research starts with the study of recommender systems namely: Collaborative Filtering, Content Based filtering, Hybrid Recommender, Demographic based filtering and Utility Based Filtering. The features and properties of recommender system are overviewed. Then we identified collaborative filtering as the most promising approach towards recommender systems. Collaborative filtering is an approach to make the recommendations based on the user and other persons past behavior and predict the items on basis of their interest. Collaborative filtering is further classified as memory based filtering, model Based filtering and hybrid filtering. Various application areas of collaborative filtering include: Books, Social Networking Sites, Movies, Music, Images, and Shopping etc. (Figure 1).

Figure 1.

Research roadmap

Best recommender system used collaborative filtering based on Data Mining approach. Data mining is the process of discovering interesting knowledge from large amounts of data stored in databases, data warehouses, or other information repositories” (Han & Kamber, 2001). Finally, authors come up with a broad classification framework showing the use of data mining in collaborative filtering for effective recommendations to the users. Figure 1 shows the research roadmap that has been followed in the chapter.