基于改进用户属性评分的协同过滤算法

为解决在基于用户的推荐算法中,用户相似度计算精度较低、缺乏个性化等问题,提出一种基于改进用户属性评分的协同过滤算法(IUAS-CF)。针对个性用户、偏执用户等在评分矩阵上存在的评价值范围差异,基于现有的相似度计算公式设计一种适应于计算个性化用户相似度的距离度量公式;针对用户自身存在影响用户抉择的用户属性,设法将用户属性评分量化,将其引入相似度计算公式中。实验结果表明,IUAS-CF算法能更真实地反映用户评分偏好,提高了推荐系统的推荐精度,更好地满足了用户对系统的个性化需求。

Collaborative filtering algorithm based on improved user attribute score

To solve the problems of low accuracy and lack of personalization in user-based recommendation algorithm,a collaborative filtering algorithm based on improved user attribute score(IUAS-CF)was proposed.Aiming at the difference of evaluation value range between individual users and paranoid users on the score matrix,a distance measurement formula suitable for calculating the similarity of individual users was designed based on the existing similarity calculation formula.Aiming at the user attributes affecting the user’s choice,the user attributes score was quantified and it was introduced into the similarity calculation formula.Experimental results show that the proposed IUAS-CF algorithm can more truly reflect users’rating preferences,improve the recommendation accuracy of the recommendation system,and better meet users’personalized needs of the system.