首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于Item-Based协同过滤农业高校图书个性化推荐算法研究——以河北农业大学为例
引用本文:赵鹏博,韩宪忠,王克俭.基于Item-Based协同过滤农业高校图书个性化推荐算法研究——以河北农业大学为例[J].湖北农业科学,2017,56(6).
作者姓名:赵鹏博  韩宪忠  王克俭
作者单位:河北农业大学信息科学与技术学院,河北 保定,071001
基金项目:河北农业大学理工基金重点项目,河北省高等学校科学技术研究项目
摘    要:针对高校用户对图书的个性化需求,运用用户对图书的评分,构建了基于Hadoop和Mahout的图书推荐系统。通过Hadoop中分布式文件系统(HDFS)和Map/Reduce计算模型的应用,发现当Hadoop中节点数不断增加时,计算时间不断减少,实时响应效率得到了提高;通过对Mahout中传统的Item-Based聚类协同过滤推荐算法进行改进,利用MAE值对传统和改进后的协同过滤算法进行比较,发现图书推荐的精度进一步提高。总体来说,推荐系统改善了传统单机运行内存严重不足和推荐结果不精确的问题。

关 键 词:高校图书馆  个性化推荐算法  协同过滤算法  大数据

Research on Algorithm of Personalized Recommendation Books in Agricultural University Based on Item-Based Collaborative Filtering:Taking Agricultural University of Hebei as An Example
ZHAO Peng-bo,HAN Xian-zhong,WANG Ke-jian.Research on Algorithm of Personalized Recommendation Books in Agricultural University Based on Item-Based Collaborative Filtering:Taking Agricultural University of Hebei as An Example[J].Hubei Agricultural Sciences,2017,56(6).
Authors:ZHAO Peng-bo  HAN Xian-zhong  WANG Ke-jian
Abstract:For personalized book needs of users at colleges and universities, we build a book recommendation system based on Hadoop and Mahout using the scores given by users. Through the application of Hadoop distributed file system (HDFS) and Map∕Reduce calculation model, we found that the calculation time is reduced while real-time response efficiency is im-proved with the increasing of the number of notes in the Hadoop. Through the improvement of traditional Item-Based collabo-rative filtering recommendation algorithm, which is based on item clustering in the Mahout, we compare the traditional collab-orative filtering algorithms with the improved collaborative filtering algorithms by using MAE, and find that the precision of the recommendation is further improved. In general, this experiment improves the problem of out-of-memory for the running of traditional single machine and the inaccurate results of recommendation.
Keywords:university library  personalized recommendation algorithm  collaborative filtering algorithm  big data
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号