摘要
针对个性化服务研究中用户兴趣度估计的要求,分析用户行为特征与兴趣度的相互关系,选取页面关注时间、滚动/翻页次数、页面大小作为用户兴趣度的判别依据,提出一种基于RBF神经网络模型的用户兴趣度量化估计方法。仿真实验证明,与多元线性回归模型的计算结果相比,该方法在平均残差和预测准确度方面均有更好的效果。
The Aiming at the requirements for estimating the user interest degree in the personalization services researching,this paper analyzes the relationship between the characteristics of user＇s behavior and the interest degree and selects the page-concerned time,number of page rolling and page size as the implicit indicators,then proposes a method based on RBF neural network to quantitatively estimate the user interest degree.Experimental results show that compared with the multiple linear regression model,this method achieves better results both on the average residuals and the predicting accuracy.
出处
《计算机工程》
CAS
CSCD
北大核心
2011年第7期187-189,192,共4页
Computer Engineering
基金
国家科技支撑计划基金资助项目(2008BAH28B04)
关键词
个性化服务
用户兴趣度
RBF神经网络
personalization service
user interest degree
RBF neural network
