A formal model for user preference
Sung Young Jung, Jeong-Hee Hong, Taek‐Soo Kim · 2003
Personalization and recommendation systems require a formalized model for user preference. We present the formal model of preference including positive preference and negative preference. For rare events, we apply the probability of random occurrence in order to reduce noise effects caused by data sparseness. Pareto distribution is adopted for the random occurrence probability. We also present the method for combining information of joint feature variables in different sizes by dynamic weighting using random occurrence probability.