摘要
在电子商务网站中能够了解用户的兴趣爱好,并能恰当地推荐出用户需要的信息是一项非常重要的工作。笔者提出了一种基于web日志并结合页面特征内容及用户浏览时间的面向个性化信息推荐的用户兴趣计算方法。该方法在将网页根据特征内容分为多个主题类型的基础上,采用一种"一次会话,单一目标"的原则,可以科学的计算出用户对各主题类型和商家的兴趣度。
Knowing user＇s interest and recommending information what they need exactly is an important task in e -business websites. Author proposes a personalized information recommendation algorithm： basing on web logs, referring to the contents of web page and the time used in browsing pages. Following the principle of ＂one -session, single -target＂, this algorithm can get user＇s interest types of information and merchant scientifically by analyzing the web page which is classified by its feature contents.
出处
《微处理机》
2010年第4期86-90,共5页
Microprocessors
基金
国家科技支撑计划项目(0216002343012)
