一种改进Slope One的学习推荐算法  

An improved Slope One learning recommendation algorithm

作　　者：刘毓[1] 何锐[1] 

机构地区：[1]西安邮电大学通信与信息工程学院,陕西西安710121

出　　处：《西安邮电大学学报》2017年第4期105-108,共4页Journal of Xi’an University of Posts and Telecommunications

基　　金：陕西省工业科技攻关项目(2016GY-113)

摘　　要：针对协同过滤算法的推荐精度不足问题,提出一种改进的Slope One算法。以基于用户协同过滤算法为前提,使用皮尔逊相似性计算用户间相似度,利用Top-N方法对相似用户进行筛选,把最相似用户作为邻居集,再结合加权Slope One算法,预测项目评分,实现对用户个性化精准推荐。实验结果表明,在数据稀疏的条件下,改进算法的预测精确度优于基于用户的协同过滤算法和Slope One算法,提高了推荐质量。An improved Slope One algorithm is presented to overcome the shortcomings of traditional collaborative filtering algorithm. Based on the user collaborative filtering algorithm, the similarity between users is calculated by using Pearson？s similarity, and the similar users are selected by Top-N method. The most similar users are then used as neighbour sets and combined with weighted Slope One algorithm to achieve personalized recommendations for personalized users. Experimental results show that the prediction accuracy of the improved algorithm is better than that of the user-based collaborative filtering algorithm and the Slope One algorithm under the condition of sparse data. Thus it can improve the recommendation quality.

关 键 词：协同过滤算法 SLOPE One算法 数据稀疏 

分 类 号：TP391.3[自动化与计算机技术—计算机应用技术]