Towards Sustainable Society on Ubiquitous Networks

The massive growth of the Internet has made an enormous amount of infor- tion available to us. However, it is becoming very difficult for users to acquire an - plicable one. Therefore, some techniques such as information filtering have been - troduced to address this issue. Recommender systems filter information that is useful to a user from a large amount of information. Many e-commerce sites use rec- mender systems to filter specific information that users want out of an overload of - formation [2]. For example, Amazon. com is a good example of the success of - commender systems [1]. Over the past several years, a considerable amount of research has been conducted on recommendation systems. In general, the usefulness of the recommendation is measured based on its accuracy [3]. Although a high - commendation accuracy can indicate a user's favorite items, there is a fault in that - ly similar items will be recommended. Several studies have reported that users might not be satisfied with a recommendation even though it exhibits high recommendation accuracy [4]. For this reason, we consider that a recommendation having only accuracy is - satisfactory. The serendipity of a recommendation is an important element when c- sidering a user's long-term profits. A recommendation that brings serendipity to users would solve the problem of "e;user weariness"e; and would lead to exploitation of users' tastes. The viewpoint of the diversity of the recommendation as well as its accuracy should be required for future recommender systems.