摘要
针对读者个性化服务质量的问题,提出了基于独立分量分析的图书馆个性化及时服务系统;简要给出了系统总体和个性化服务关键子系统的体系结构;重点讨论了利用独立分量分析方法改善系统个性化服务质量的问题。实验结果显示,利用此方法进行图书分类和读者群体聚类的效果较理想;读者群体和图书类别之间可产生随读者个性或图书主题变化而变化的动态映射关系,这种映射关系可以改善个性化服务质量。
A system of personalized dynamic services in digital library based on Independent Component Analysis （ICA） was proposed and discussed in detail in order to solve the qualities of personalized services to readers. Otherwise, architectures of general personalized dynamic services system and the most important subsystem were introduced. The experiments show that better effects of books classification and readers group could be respectively achieved by using ICA, and a dynamic mapping relation that improved the qualities of personalized services, furthermore, could be generated along with the changes between readers＇ personal interests and books＇ themes.
出处
《计算机应用研究》
CSCD
北大核心
2007年第3期187-190,共4页
Application Research of Computers
基金
国家自然科学基金资助项目(60471055)
国家"863"计划资助项目(2005AA114030)
关键词
独立分量分析
文本分类
数字图书馆
个性化服务
及时服务
independent component analysis （ICA）
text classification
digital library
personalized services
dynamic services
