摘要
【目的/意义】数字期刊阅读在移动网络的推动下已成为互联网使用群体必不可少的活动,分析数字期刊服务中的用户行为能够掌握用户阅读倾向,提高推送精确度。从用户行为感知的角度出发根据用户信息分析用户行为轨迹,搭建推送模型,结合推荐算法和用户行为预测模型,实现数字期刊的用户个性化推荐。【方法/过程】本文使用聚类算法和加权二部图算法计算用户相似度、寻找用户邻居簇,进而预测用户对数字期刊的评分,基于加权Markov模型预测用户模型,生成数字期刊推送结果。【结果/结论】利用用户行为感知技术对数字期刊服务推送提出了基本的研究理论框架,依照推荐算法设计对不同用户进行个性化推荐,增强用户信赖感,提高用户对数字期刊阅读的满意度。
【Purpose/significance】Digital periodical reading has become an indispensable activity for Internet users under the impetus of mobile networks.Analyzing user behaviors in digital periodical services can help users grasp reading trends and improve delivery accuracy.This article analyzes user behavior trajectory according to user behavior information,builds a push model,and combines recommendation algorithm and user behavior prediction model,so as to achieve the purpose of personalized recommendation of digital periodical users.【Method/process】In this paper,clustering algorithm and weighted bipartite graph algorithm are used to calculate user similarity,find user neighbor clusters,and then predict users’scores on digital journals.Then,the user model is predicted based on weighted Markov model,and digital journal push results are generated.【Result/conclusion】Utilizing user behavior awareness technology,this paper proposes a basic theoretical framework for digital periodical service push.According to the recommended algorithm design,it can personalize recommendations for different users,enhance user trust,and improve user satisfaction with digital periodical reading.
作者
张继东
王蓉
ZHANG Ji-dong;WANG Rong(School of Economics and Management,Hubei University of Technology,Wuhan 430068,China)
出处
《情报科学》
CSSCI
北大核心
2019年第5期19-24,共6页
Information Science
基金
国家自然科学基金面上项目"移动社交网络环境下基于情景化偏好的用户行为感知与自适应建模研究"(71573073)
作者简介
张继东(1976-),男,山东安丘人,教授,博士,主要从事移动互联网知识管理技术、数字图书馆技术研究.
