摘要
在基于用户浏览历史的个性化服务中,网页特征提取和兴趣建模通常是基于传统全局词典进行的,但是传统全局词典容易向网页特征里引入较多的噪声数据。因此有必要采用个性化词典替换传统全局词典。针对目前个性化词典的建立无法自动获取用户兴趣网页以及缺乏网页预处理的缺点,介绍了一种基于兴趣网页的个性化词典UPDBIWP,其特点是引入基于浏览行为量化分析的兴趣网页自动捕获技术和基于超链接的网页正文提取技术,使个性化词典的建立更加智能化和自动化。通过实验验证UPDBIWP对用户的兴趣点和兴趣偏好的描述更准确。
In the peronalization services based on the browsing history of users,extraction of web pages feature and user interests modeling usually proceed based on the global dictionary.But noisy data may be injected into web pages features by the global dic tionary easily.So it's necessary for user personalization to substitute for global dictionary.An user personalization based on inter esting web pages,in which the technology of extraction of user interesting web pages based on the browsing action quantitative analysis and body of web pages based on hyperlink is introduced for establishing of user personalization dictionary more automati cally and intelligently,has been introduced for the user personalization dictionary presently,which is lack of pretreatment for web pages and unable to acquire interesting web pages automatically when establish.The user personalization dictionary based on in teresting web pages could descripte the interests and preference of interests of users more accuracy.
出处
《电脑知识与技术（过刊）》
2012年第10X期6992-6995,共4页
Computer Knowledge and Technology
基金
重庆市教委科技基金项目(KJ091306)
关键词
兴趣网页
超链接
个性化词典
网页特征提取
兴趣建模
interesting web pages
hyperlink
user personalization dictionary
extraction of web pages feature
user interests mod eling
