摘要
电子邮箱的推送功能对信息传播和推广有重要作用,而科研在线云平台的邮箱推送功能出现信息过载问题后却使得推送邮件可能变成垃圾邮件。为解决该问题,本文介绍了一种基于科研在线云平台的个性化推荐系统,通过推荐模型将资源进行排序,并按周选取得分最高的一部分资源通过推送邮件推荐给用户。系统运行测试表明,通过个性化推荐的动态信息更加符合用户兴趣,有助于增加科研在线团队文档库动态汇总邮件的点击率,增强用户体验。
Push E-mail function plays an important role in information dissemination and promotion, however, with the emergence of information overload, the push email functionality of the Research Online cloud plateform may make the push mail produce spam.To solve this problem, in this paper we personalized one kind cross-application recommendation system based on the Research Online cloud platform. The recommended system sorts resources by recommended models, and recommends to the users those are selected from the highest scoring part of the weekly resource. System tests show that the dynamic information recommended information by the personalized system is more in line with the user＇s interest. This will help increase the dynamic summary of the document library mail CTR and enhance the user experience.
出处
《科研信息化技术与应用》
2016年第1期76-83,共8页
E-science Technology & Application
关键词
个性化推荐
跨应用推荐
IFR模型
科研在线
personalized recommendation
cross-Application recommendation
IFR model
the Research Online
