A Novel Weight for Recommendation: Item Quality
Sunghoon Cho, Moohun Lee, Jeongseok Kim, Bonghoi Kim, Euiin Choi · 2008
The many researches of recommendation technique already have been performed. These techniques were used mostly content-based filtering, collaborative filtering or hybrid filtering approach. They are performed the recommendation using userpsilas rating, user similarity, itempsilas features, user profiles. They consider several features of user and user profiles but they do not consider a quality of items itself. Because they donpsilat be able to efficiently define the quality definition of all item and they donpsilat be able to easily get user feedbacks. To solve the difficulty of the measurement of item quality and apply item quality as a weight to recommendation techniques, we measure item quality through popularity of item and user awareness. And we propose an approach to apply item quality to recommendation.