Combining Dynamic Agents and Collaborative Filtering without Sparsity Rating Problem for Better Recommendation Quality.
Saranya Maneeroj, Hideaki Kanai, Katsuya Hakozaki · 2001
Information Filtering and Collaborative Filtering techniques have been used for selecting information based on the user’s previous preference tendency and the opinion of other people who have similar tastes with the user. Combining both Information Filtering and Collaborative Filtering, or a hybrid systems, have also been proposed to get better recommendation results. In this paper, we present an improved recommendation method that copes with the sparsity problem of the hybrid systems and increases the accuracy of recommendation results. We also present an experimental recommender system for movie, called e-Yawara (extended Yawara), for implementing our method. Evaluation shows that e-Yawara is more efficient and provides more satisfactory result than conventional filtering system.