论文部分内容阅读
为使用户在手持设备上提取用户兴趣是重要基础,提出一种基于用户隐式反馈的方法。该方法从网页中提取兴趣块(用户感兴趣的网页内容),假设用户的浏览行为如在块上的停留时间、滚动次数、滚动速度、进入链接的次数等与用户是否对块感兴趣相关。从手持设备上的浏览行为中提取了34种隐式反馈特征,通过分析18个用户在608个网页的9474个块上的浏览行为,验证了其中29种特征与兴趣块的相关性,并且这种相关性能用于跨用户、跨网站的兴趣块提取。该文的研究成果可用于建立个性化的用户偏好模型,应用在手持设备上的自适应网页浏览中。
In order to make the user extract the user interest on the handheld device is an important foundation, a method based on user implicit feedback is proposed. This method extracts the interest block (the content of the webpage that the user is interested in) from the webpage. It is assumed that the user’s browsing behavior such as the staying time on the block, the number of scrolling times, the speed of scrolling, the number of times of entering the link and the like are related to whether the user is interested in the block. 34 implicit feedback features were extracted from the browsing behavior on the handheld device. By analyzing the browsing behavior of 18 users on 9474 blocks of 608 web pages, the correlation between 29 features and the interest blocks was verified, and The correlation performance is used to extract interest blocks across users and across websites. The research results of this paper can be used to build a personalized user preference model for adaptive web browsing on handheld devices.