An Online Behavioral Advertising Framework To Protect Users’ Clickstreams

In the past10years, the network has become an essential part of the human in theworld. Compared with traditional advertising, the Internet became an important partin the advertising industry and gradually become a major advertising channel. Thereare two necessary concepts in online behavioral advertising–Tracking and targeting.Because of the accurate targeting effect for the users, more and more advertisingpublishers choose online behavioral advertising (OBA). However, in order to targetthe user’s preferences accurately, the advertising network (such as Google, yahoo) hasto collect the user’s network click-stream information (streams of URLs of Web pagesvisited). In this process, the advertising network does not get the user’s permission togain their private information hidden in the click-streams. According to our survey,with the progressive emphasis on security privacy, people began to challenge and tryto stop the tracking behavior from the advertising network. Therefore, how to protectthe user’s privacy while maintaining the targeting effect of online behavioraladvertising framework became the popular problem in privacy security. Today, sometechnical solutions have been proposed to prevent advertising networks to track theuser. However, these solutions are radical and they deny and abandon the role of adnetwork to target users’ interests.In this paper, we propose a new OBA framework—acceptable tracking (AT) tosolve leakage problem in the traditional online behavioral advertising model. On onehand, AT largely retained the role of targeting function by the advertising network. Onthe other hand, it will protect the user’s click-stream information. AT reserved onlinebehavioral advertising business model and it doesn’t need to install the Web browserplug-in to prevent ad network tracking user. Therefore, AT is a compromise solutionfor OBA industry.This article describes the design details and the promotion of AT, and verifyaccuracy of AT profiling. At the same time, we have completed the Web browser performance analysis of AT and proved that network traffic burden is acceptable.Finally, we compared AT with the existing solutions to deal with the problem ofprivacy leakage in OBAand We foune that AT has the following advantages.First, AT does not require the OBA parties to make huge changes, so thefeasibility of AT greatly enhanced. Secondly, although we introduced protectionmechanisms, AT maintain the accuracy of targeting. Thirdly, AT makes the revenueprocess transparent and auditable. Finally, different from other solutions that buildin-browser user profiles, AT completely detect click fraud.