Research Of Precautionary Technique On Click Fraud In Online Advertising

With the development of technology and the advancement of media, traditionaladvertising business model can not meet the need of development. As a new media platform,the Internet plays a more important role in the market. Because of the convenience andtimeliness of internet information, network platform market has also been widely used invarious fields, and more and more businessmen adopt online advertising to promote theirsales. The payment models of advertising attract much attention, among which pay-per-clickmodel is a popular and simple way now. The pay-per-click model brings the convenience ofthe commercial activities and the click fraud danger simultaneously. The appearance of clickfraud is a serious threat to the healthy development of the network market. So how to preventand detect click fraud on online advertising is the key issue.In online advertising, click fraud occurs mainly due to the illegal market competition orin order to obtain excessive benefits. Recently, the precautionary technique researches onclick fraud are mainly prevention technique and detection technique. In this thesis, we analyzethe state of the art of several classical click fraud identification techniques. Then, we utilizethe density-based anomaly detection algorithm to characterize users’ behavior on the network.Finally, the game mode was founded to analyze the behavior character of the both seller andthe buyer to prevent the click fraud. The research contents are as follows:(1) The pay-per-click mechanisms and characteristics, the reasons for the appearance ofclick fraud phenomenon based on the pay-per-click model characteristics are described. Thepay-per-click model is different from other models for each click will produce a pay forcustomers and therefore attracts more and more clicks which make use of this mechanism toearn money.(2) The density-based anomaly detection algorithm for click fraud detection is proposedby studying the several current existing precautionary techniques on click fraud.(3) By adopting game theory, we put forward the click fraud prevention technique basedon user behavior.(4) From the precision and recall point of view, we deploy the experiments on thebenchmark dataset. The experiment results illustrate the effectiveness and efficiency providedby our precautionary technique on click fraud.