摘要
协同过滤推荐算法是电子商务系统的关键技术,为了解决当前协同过滤推荐算法中存在的错误大、速度慢等缺陷,以获得更优的协同过滤推荐效果,设计了综合用户属性和相似度的协同过滤推荐算法。首先分析当前电子商务系统中的协同过滤推荐算法研究现状,找到各种协同过滤推荐算法的局限性,然后综合考虑用户个性化特征建立用户属性评价矩阵,并根据评价矩阵计算了用户之间的相似度,根据相似度实现商品推荐,最后采用Java编程实现协同过滤推荐程序,并采用具体数据进行了协同过滤推荐仿真测试。这个算法减少了协同过滤推荐时间,协同过滤推荐速度得到了明显加快,降低了协同过滤推荐误差,协同过滤推荐精度要远高于当其它协同过滤推荐算法,具有很好的实际应用价值。
Collaborative filtering recommendation algorithm is the key technology of e-commerce system.In order to solve the shortcomings of current collaborative filtering recommendation algorithms,such as large error and slow speed,and to obtain better collaborative filtering recommendation effect,a collaborative filtering recommendation algorithm integrating user attributes and similarity is designed.Firstly,the current research status of collaborative filtering recommendation algorithms in e-commerce system is analyzed,and the limitations of various collaborative filtering recommendation algorithms are found.Secondly,the user attribute evaluation matrix is established by considering user’s personalized characteristics.Then,the similarity between users is calculated according to the evaluation matrix,and the commodity recommendation is realized according to the similarity.Finally,Java programming is used to realize the recommendation degree of collaborative filtering,and the simulation test of collaborative filtering recommendation is carried out with specific data.The proposed algorithm speeds up the collaborative filtering recommendation,speeds up the collaborative filtering recommendation speed and reduces the collaborative filtering recommendation error.The collaborative filtering recommendation accuracy is much higher than other collaborative filtering recommendation algorithms,and has better practical application value.
作者
农艺
唐忠
NONG Yi;TANG Zhong(School of Information and Management,Guangxi Medical University,Nanning 530021)
出处
《微型电脑应用》
2019年第11期27-29,共3页
Microcomputer Applications
基金
广西云计算与大数据协同创新中心研究课题(YF16201)
关键词
电子商务系统
用户个体性特征
数据挖掘
信任度评价
推荐算法
E-commerce system
User personality characteristics
Data mining
Trust evaluation
Recommendation algorithms
