一种新的数字家庭个性化服务推荐方法

An Innovative Digital Home Personalized Service Recommendation Method

针对数字家庭系统用户找到满意服务资源越来越困难的问题,提出了使用协同过滤算法进行个性化推荐的方法。针对传统算法中用户评分数据存在的稀疏性的问题,提出了一种改进的基于用户行为数据的协同过滤推荐算法。通过引入用户使用资源的具体时间、总次数和时间长度这3个维度构建用户兴趣模型,利用该模型寻找目标用户的最近邻居。实验证明,该方法可以有效地为用户提供个性化推荐,进一步扩展了协同过滤方法的应用范围。

It is increasingly difficult for digital home system users to find satisfactory service resources.To solve the problem,a personalized recommendation was proposed by using collaborative filtering algorithm.Given that the sparse user rating data of traditional algorithm,it came up with an improved collaborative filtering algorithm based on the user behavior data.A user interest model was built from three dimensions,including the specific time,total frequency and time length,and then the interest model was used to find the nearest neighbor of target user.The experiment proves that the method is effective to provide personalized recommendation for digital home system users.The method extends the application range of collaborative filtering method.