【摘要】 协同过滤技术(Collaborative Filtering)成功应用于个性化推荐系统中,为了用户能更准确地获取信息,提出了利用领域知识进行相似度计算的协同过滤算法,利用用户兴趣与领域本体中概念的映射关系,构建用户兴趣本体,发掘用户兴趣模式,通过融合评分项目相似度和用户相似度的计算,使用户在评分的共同项目很少或为零的情况下也能找到最近邻进行协同推荐。将这一方法应用到某健康系统中进行实验分析,该方法不仅解决了传统的基于项目的协同过滤带来的问题,而且还提高了推荐系统的推荐质量。更多还原

【Abstract】 Collaborative filtering techniques are currently being successfully used in personalized recommendation system,in order to obtain more accurate information for users,this paper gave the collaborative filtering algorithms by the use of domain knowledge for similarity calculation,utilizing the mapping of domain ontology and the concept in user interest ontology,constructed users interest ontology,explored user interest model,through the calculation of integratiing similarity of rating items and user similarity,users could find the nearest neighbor to give a collaborative recommended under the situation of little or zero joint rating items.This method is applied to an experimental analysis of the health system,the method not only solves the problems traditional item-based collaborative filtering brought about,but also to improve the recommendation quality of the recommendation system.更多还原