在网络广告业中出现的欺诈点击行为,使得搜索引擎企业以及广告主的利益受到了严重损害,致使点击付费模式遭到质疑,欺诈点击已经成为阻碍网络广告业健康发展的一大顽疾.针对网络广告业发展所面临的此种困境,提出一种基于用户行为分析的广告欺诈点击检测技术.首先创建用户行为数据仓库,然后运用贝叶斯分类方法对用户行为数据进行点击合法等级预测,最后结合博弈控制机制对用户点击有效性进行最终判断.%At present, the behaviour of frauduient clicks in internet advertisements severely damages the interesis of the search engine business and the advertisers, which leads to the payment pattern of per-clicking to be questioned. Fraudulent clieks have become a persistent ailment against the healthy development of internet adverisement. Aiming at the predicament encountered by the internet advertising industry,a technology of fraudulent clicka detetion on adverisement based on user behaviour analysis is proposed. After a user behaviour data warehouse is built, the legal grades of user clieking are predicted by using Bayesian classifier based on user behaviour data. Finally, the validity of user clicks is ultimately judged in conjunction with game control mechanism.
展开▼
