摘要
针对传统的Hadoop MapReduce框架下数据计算效率低下的问题,选用基于内存迭代计算的Spark框架,提出融合用户偏好与上下文信息的加权矩阵分解算法,设计并实现了针对目标用户的个性化地点推荐系统.结果表明,系统的整体架构分为数据源、数据仓库、核心业务与数据展示4个模块,模块间的稳定传输保证了推荐系统的顺利运行.在真实数据集的基础上对系统进行了实验测试,验证了设计系统的高准确率与高召回率.
To address the problem of inefficiency of data computation under traditional Hadoop MapReduce framework,the Spark framework based on memory iterative computation was selected,the weighted matrix decomposition algorithm combining user preference and context information was proposed and a personalized place recommendation system for target users was designed and implemented.The results show that the overall architecture of the system is divided into four modules:data source,data warehouse,core business and data display.The stable transmission between modules ensures the smooth operation of the recommendation system.,The experimental test of the system was carried out on the basis of the real data set to verify the high accuracy and recall rate of the designed system.
作者
黄东
陈光
李海滨
杨朔
HUANG Dong;CHEN Guang;LI Haibin;YANG Shuo(College of Information and Control Engineering,Qingdao University of Technology,Qingdao 266520,China;Beidou Navigation Laboratory,Qingdao Optoelectronic Engineering Technology Research Institute,Qingdao 266011,China)
出处
《辽宁工程技术大学学报（自然科学版）》
CAS
北大核心
2020年第6期533-540,共8页
Journal of Liaoning Technical University (Natural Science)
关键词
推荐系统
Spark框架
加权矩阵分解
个性化地点推荐
迭代计算
相似计算
recommendation system
spark framework
weighted matrix decomposition
personalized location recommendation
iterative computations
similarity calculation
