| 摘 要: 随着互联网及物流运输行业的快速发展，越来越多的人选择在网上挑选服饰类商品。基于服饰类商品具有 重复购买率低、搭配性强、受当季流行因素影响大等特点，提出了一种基于协同过滤与专家推荐的混合推荐策略，在为 商品引入流行因子的基础之上，为用户提供了一种更为个性化、时尚化的推荐结果。由于业务系统涵盖了海量的商品及 用户数据，单机计算系统难以满足推荐系统对计算资源的需求，在基于Hadoop平台的基础之上，构建了一套离线分布 式推荐系统，为解决大数据应用背景下的数据计算问题提供了可行性案例。 |
| 关键词: 推荐系统；专家推荐；协同过滤；Hadoop；分布式计算 |
|
中图分类号: TP311.5
文献标识码: A
|
|
| The Design and Implementation of E-commerce Recommended System Based on Experts Recommend |
|
GUO Qing，SUN Jian
|
( College of Information Science and Technology, Beijing University of Chemical Technology, Beijing 100029, China)
|
| Abstract: With the rapid development of the internet and the logistics industry,more and more people choose to buy clothes on the internet.Base on the features of apparel goods,such as the low repurchase rate,the high matching requirements,the high dependence on seasonal fashion,the paper proposes a mixed recommendation strategy based on collaborative filtering and expert recommending.The new strategy recommends users with more personalized and more fashionable results by introducing fashion elements in the system.Because of the massive data of goods and users covered in the system,stand-alone operating systems are apparently unable to meet the requirements.The paper constructs a distributed offline computing system based on the Hadoop platform,which provides a feasible case of computing in the application of big data. |
| Keywords: the recommendation system;experts recommending;collaborative filtering;distributed computing;Hadoop;the mixed recommendation strategy |
