Issue No. 03 - May/June (2007 vol. 22)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/MIS.2007.48
Jinhyung Cho , Seoul National University
Kwiseok Kwon , Seoul National University
Yongtae Park , Seoul National University
With the proliferation of e-commerce on the Web, e-commerce providers will need to offer recommender systems if they wish to remain competitive. One of the most successful recommendation methods is collaborative filtering. To provide recommendations, conventional CF methods use only a single recommender group (that is, a single information source). Consequently, they have several limitations that make them unsuitable for high-involvement, knowledge-intensive product domains such as e-learning. A new CF method, based on group behavior theory from consumer psychology, attempts to overcome these limitations. To adapt CF to Web-based e-learning content services, this method forms dual recommender groups: similar users and expert users. In experiments, a recommender system employing this method outperformed conventional CF methods in situations involving variations in the product domain and in data sparsity. This article is part of a special issue on recommender systems.
recommender system, collaborative filtering, information source, group influence, e-learning content
Y. Park, J. Cho and K. Kwon, "Collaborative Filtering Using Dual Information Sources," in IEEE Intelligent Systems, vol. 22, no. , pp. 30-38, 2007.