摘要
通过分析目前推荐技术在电子商务系统中的应用优势,并针对当前产品交易系统的无评分、产品量大和难以分类等现状与问题,设计了一种基于用户交易行为的隐语义模型推荐算法。该算法从用户的隐式交易行为出发,采用隐语义模型推荐算法,构建用户-产品兴趣模型,并加入K均值算法划分隐式特征聚类。实验验证表明,该算法在满足用户的个性化需求的同时,可提高电子商务系统的产品推荐效率。
Through analysis of the application advantages of recommendation technology in e-commerce system,and in order to solve the present situation and problems of the product trading system without scoring,large volume of products and difficult classification,a latent factor model recommendation algorithm based on user transaction behavior was designed,which starts from the implicit user transaction behavior,and constructs the interest model between users and products,which uses latent factor model recommendation algorithm,and K-means algorithm is used to cluster implicit feature. The experimental results show that the algorithm meets the individual needs of users,and can improve the recommendation efficiency of e-commerce system.
作者
梁婧文
蒋朝惠
Liang Jingwen;Jiang Chaohui(College of Computer Science and Technology,Guizliou University,Guiyang 550025,Chin)
出处
《微型机与应用》
2017年第21期15-18,25,共5页
Microcomputer & Its Applications
基金
贵州省基础研究重大项目(黔科合JZ字[2014]2001-21)
关键词
推荐算法
用户交易行为
隐语义模型
K均值算法
recommendation algorithm
user transaction behavior
latent factor model
IK means algorithm