The collaborative filtering algorithm based on domain ontology and user preferences

This paper proposes a method of user ratings similarity, based on forgetting function, which adjusts the importance of the user ratings according to time, considering the impact of the changes in user´s preferences. While the user clustering algorithm is presented, which takes into account the personal characteristics of the user´s information to affect its form ultimately choose to purchase goods weighting factor improved characteristics of the user similarity algorithm, reducing the nearest neighbor range of options.