中文摘要：

针对传统协同过滤算法当用户项目评价矩阵呈数据稀疏状况时存在不足这一问题,提出了一种基于项目属性偏好的协同过滤算法。该算法基于用户的特定属性偏好值计算用户之间的相似度,并在此基础上进一步预测用户对项目的具体属性未评分部分的评分,以此改善原始用户项目评价矩阵的数据稀疏状况,获得稠密数据。最后采用相应的协同过滤算法实现基于项目属性偏好的个性化推荐。该方法通过实践验证,能够很好地提升项目推荐的效率。

英文摘要：

To address the shortcomings of traditional collaborative filtering algorithm for data sparsity of the user project evaluation matrix, a collaborative filtering algorithm based on the item attribute preference is proposed. The algorithm calculates the similarity between users by user preference value for project properties. Then predict item ratings that users have not rated based on user similarity to increase data density of the original user project evaluation matrix. Finally adopts the corresponding collaborative filtering algorithm based on the project properties preference to achieve the personalized recommendation. This method can improve the efficiency of recommendation.