Advancement of recommender system based on clickstream data using gradient boosting and random forest classifiers

In this article, we have made an improvement on Kim et al. (2005) approach of recommending products and further developed a novel recommender system. The proposed system analyzes the clickstream data obtained from an ecommerce site and predicts the preference values of the customer for the products clicked but not purchased using more efficient classifiers such as random forest and gradient boosting and then Collaborative Filtering is used to recommend products. In Collaborative Filtering, a better similarity measure i.e. Proximity Significance Singularity along with efficient clustering algorithm i.e. rough set clustering algorithm is used which helps in making better recommendations. To determine the effectiveness of the proposed approach, an experimental evaluation have been done which clearly depicts the better performance of recommender system as compared to Kim et al. (2005).