基于模糊处理和上下文敏感SVM的协同过滤推荐新方法

协同过滤系统是电子商务最重要的技术之一,用户相似度算法的优劣直接决定推荐性能的准确程度。现有推荐方法忽略用户上下文特征,因而用户相似度判定较差。针对该问题,该文提出了FSVMCF方法,该方法采取模糊上下文数据及上下文敏感SVM和协同过滤相结合方法,提高了推荐准确度。该文实验以汽车、飞机、火车等交通工具图像作为推荐对象,实验结果验证了该方法对于图像推荐的性能有较大提高。

A New Image Recommendation Algorithm Based on Fuzzy and Context-aware SVM

Recommendation system is one of the most important technologies applied in e-commerce.The Similarity measuring method is fundamental to recommendation system's performance.The Traditional methods ignore the context,so it brings poor similarity result.Based on fuzzy and context-aware SVM,a novel similarity measuring method FSVMCF has proposed to solve this problem.By using the images containing car,plane and train.the experimental results show that this method can efficiently improve the recommendation syste...