融合动态K近邻Slope_One的协同过滤推荐算法

Integrating Dynamic K-nearest Neighbor Slope_One into Collaborative Filtering Algorithm

传统协同过滤推荐算法存在数据稀疏的问题,这会导致算法精确度不足。Slope_One算法简单高效,可以预测用户对某个物品的评分。因此,论文提出融合动态K近邻Slope_One的协同过滤推荐算法,提高推荐算法的精确度。首先利用改进余弦相似度公式计算用户相似度,筛选出K个近邻用户进行平均评分偏差计算,利用Slope_One算法预测相应的用户评分并对评分矩阵进行有效填充,然后在新的评分矩阵上,利用基于物品的协同过滤算法进行推荐。

Data sparse is a problem of traditional collaborative filtering algorithm,which will cause the algorithm to be insuffi-cient.The Slope_One algorithm is simple and efficient,and can predict the user's rating of an item.Therefore,this paper proposes a collaborative filtering recommendation algorithm combining dynamic K-nearest neighbor Slope_One to improve the accuracy of the algorithm.First,the improved cosine similarity formula is used to calculate the user similarity,K neighbor users are screened to cal-culate the average score deviation,the Slope_One algorithm is used to predict the corresponding user score,and effectively the score is filled into data matrix,and then the item-based collaborative filtering algorithm is used for recommendation.