摘要
互联网使网上购物成为一种新的生活方式,但同时网络信息超载不可避免。提高电子商务的推荐精度已经成为各电商的重点。首先对现有推荐系统介绍分析并找出不足,提出结合眼动追踪和鼠标行为改进推荐系统的假设。利用AHP层次分析法分析用户感兴趣的眼动参数权重,分析用户感兴趣程度伴随着用户眼部行为的变化。再通过马尔科夫链来预测用户鼠标行为。最后通过鼠标行为分析辅助眼动参数分析用户真正兴趣,为判别用户真正兴趣提供依据。
Online shopping is a new style of life which brought by internet. The over information overload can not be avoided. Improving the accuracy of recommend has become the focus of attention of the online retailers. In this paper,the author firstly introduced and analysised the defect of the existing recommendation system and proposed an assumption that combining eye tracking and mouse actions can improve the recommendations system. Using AHP analytic hierarchy process to analyze the weight of eye movement parameters,Analysis of the degree of user interest with the changes in the user’s eye behavior. Through the markoff chain to predict user’s mouse behavior. Finally,through the analysis of the mouse behavior and the analysis of the eye movement parameters to analyze the user’s real interest,in order to provide a basis for judging the true interest of users.
出处
《物流工程与管理》
2016年第3期180-183,共4页
Logistics Engineering and Management
基金
上海市人才发展基金项目(201508)集装箱码头装卸系统智能调度研究