目前用户在线活动后的目标和喜好引起广泛关注。为了满足不同用户的信息需求,从娱乐角度来对用户在线活动意向进行了研究。在无用户显性反馈时,分析了由网页来获取用户的娱乐意向。并对娱乐意向进行了定义。然后提出了网页娱乐意向学习框架,通过机器学习方法支持向量机(SVM)建立娱乐意向识别模型,实现了从网页来识别用户娱乐意向目的。根据在开放目录项目(ODP)实验结果统计,娱乐意向识别能力EF可达到85.7%。
There have been recent interests in understanding goals and preferences behind a user' s online activities. In order to satisfy different users' need, the users' online activities goals are considered from the aspect of entertainment. User entertainment intention is defined and it can be gotten from web pages contents without user' s explicit feedback. Then the framework of page EI learning is presented. The model of EI identification is built using support vector machine (SVM) to detect EI from web pages. Experiments on ODP data set show that EF of entertainment intention identification can reach 85.7%.