EPF: An Elderly Personalization Features Based Collaborative Filtering Algorithm for Pension Service

How to recommend appropriate pension services to the elderly is an active research problem in health care fields in recent years. Traditional collaborative filtering(CF) algorithm does not take into account the elderly personalization features, which in turn leads to poor recommendation results. In this paper, an elderly personalization features based CF algorithm was proposed which can makes the recommended services meet the customer's personalized requirements. In the algorithm, personalized requirements form the elderly customers can be obtained by the classification of pension services. Based on classification results, the customer personalization features model can be built, in which customers are divided into several groups. Later, a personalized recommendation set is obtained according to the personalization features model. And, service similarity is calculated on the basis of customer ratings and the service category in the set. Finally the CF algorithm can be employed to detect appropriate pension services to the elderly. The experiments show that the algorithm can improve the accuracy of predicted ratings, as well as the satisfaction of elderly customers.