摘要
为了在海量书籍中快速选择,针对高校读者同一学科在不同学习阶段知识背景相似的特点,提出一种建立基于矩阵分解的隐语义模型与时间效应的融合算法,对高校图书进行个性化图书推荐。该算法运用随机梯度下降法求解用户-项目评分矩阵;针对冷启动问题提出一种改进的解决策略;通过评价指标平均绝对误差MAE与均方根误差RMSE验证该算法推荐的准确性。通过大量实验结果验证了该算法的可行性和有效性。
In order to select quickly in the mass of books,aiming at the characteristics that university readers have the same knowledge background in different stages of learning,this paper proposed a fusion algorithm based on matrix decomposition of latent factor model and time effect,and made a personalized book recommendation for university books.The algorithm firstly used the stochastic gradient descent method to solve the user-item scoring matrix. Second,an improved solution was proposed to the problem of cold start. Finally,the accuracy of the proposed algorithm was verified by the mean absolute error MAE and the root mean square error RMSE. The results of a large number of experiments verified the feasibility and effectiveness of the algorithm.
作者
李薛剑
刘梦雅
海健强
吴雪扬
余雪莉
Li Xuejian;Liu Mengya;Hai Jianqiang;Wu Xueyang;Yu Xueli(School of Computer Science and Technology, University of Science and Technology of China ,Hefei 230026 ,Anhui, China;Co-Innovation Center jbr Information Supply and Assurance Technology ,Anhui University, Hefei 230601 ,Anhui, China;Anhui Science and Technology Trade School, Bengbu 233000,Anhui, China)
出处
《计算机应用与软件》
北大核心
2018年第5期130-134,189,共6页
Computer Applications and Software
基金
安徽大学信息保障协同创新中心开放课题(Y01002454)
安徽大学大学生科研训练计划项目(201610357400)
关键词
推荐系统
隐语义模型
时间效应
冷启动
高校图书馆
Recommendation system
Latent factor model
Time effect
Cold start
University library
