A privacy-aware framework for targeted advertising
Introduction
Online advertising provides financial support for a large portion of today’s Internet ecosystem, and is displayed in a variety of forms embedded in web sites, emails, videos and so on. As the effectiveness of advertising largely depends on the relevance between the delivered advertisements (ads) and users’ interests, a popular paradigm for current online advertising system is targeted advertising, where advertisers hire an ad broker to deliver ads to potentially interested users by analyzing users’ online profiles or behaviors [1]. Targeted advertising is beneficial to both advertisers and users: advertisers can gain higher revenue by advertising to users with a strong potential to purchase, and the users in turn receive more pertinent and useful ads that match their preferences and interests. A recent survey [2] revealed that targeted advertising brings 2.68 times revenue per ad compared with non-targeted advertising. Due to the increased effectiveness and benefits, a number of advertisers around the world have already turned to online targeted advertising systems. Examples of such advertising systems include Google AdWords [3] that deliver customized ads based on search items, and Ink TAD [4] that pushes ads according to location information revealed in user’s emails.
Although targeted advertising benefits both advertisers and users, it has raised severe privacy concerns. A recent survey [5] of 2253 participates conducted in 2012 reported that the majority of respondents expressed disapproval of targeted advertising due to privacy disclosure. Such privacy threats come from the fact that ad brokers aggressively track users’ online behaviors to obtain their preferences and interests, which can be sensitive to the users. For example, the behavior of searching for a certain kind of medicine implies that the user is likely to have certain relevant disease, whose disclosure is considered as a violation of the user’s privacy. Moreover, ad brokers rarely have clear statements about how the obtained behavioral data will be used and whom the data will be shared with. Untrusted ad brokers may sell such personal information to some adversaries without the user’s permission. Being aware of such privacy risks, users are reluctant to embrace the practice of targeted advertising [6], which hinders the effectiveness of online advertising systems.
To maintain the merits brought by targeted advertising, it is essential to incentivize users to participate in such systems. Existing studies [7], [8], [9] have focused on privacy preserving mechanisms to encourage users to involve in the targeted advertising systems. These mechanisms either assume another trusted entity sitting between users and ad brokers [7], [8], or require users to send perturbed clicking information to hide users’ true data. However, these changes made on the framework of existing targeted advertising systems provide privacy protection at the cost of the benefits of ad brokers or advertisers. The ad brokers may not be in favor of introducing an extra entity to share their ad targeting duty, which is the main source of their revenue. Similarly, the advertisers may be dissatisfied with the perturbed clicking information as perturbation undermines the accuracy of click information, which normally determines their payments [10]. Without guaranteed revenue, the advertisers and ad brokers naturally tend to maintain the adoption of traditional targeted advertising systems instead of upgrading the systems to provide privacy protection. This conflict between users and advertisers/ad brokers hinders the adoption of privacy-aware mechanisms in advertising. To promote the adoption of the privacy-aware advertising systems, the interests of all entities should be guaranteed, which, unfortunately has not yet been addressed by existing proposals.
In this paper, we propose a privacy-aware framework to boost the adoption of privacy preserving targeted advertising systems. Users, the ad broker and advertisers are assumed to be rational and selfish entities, who care only about their own interests. To ensure the interests of all entities, our framework introduces an economic compensation mechanism for privacy leakage. Such economic compensation for privacy loss has already been widely considered in the literature [11], [12]. Besides, many companies, including Bynamite, Yahoo, and Google, are also engaging in the purchase of users’ private information in exchange for monetary or non-monetary compensation [13], [14], [12]. Under the proposed framework, the ad brokers compensate economically for the users’ privacy leakage in order to incentivize users to click their interested ads. On the one hand, the users, with the expectation of receiving compensation, are inclined to click ads of interests. On the other hand, as the compensation can improve click-through rate and bring the ad broker more revenue, the ad broker is willing to provide certain amount of compensation for users whose ad clicks reveal their private interests. However, to support this framework, there are still several questions that need to be answered. First and foremost, in order to compensate privacy loss, it is essential to quantify privacy information leakage in ad clicks. Second, how much compensation should be provided for each user? The more compensation provided, the more inclined users are to click ads; while the ad broker pays more for the users’ privacy loss. Moreover, how should advertisers pay the ad broker for the ad clicks they benefit from? The amount of payment to the ad broker has an impact on the privacy loss compensation allocated to users, which in turn affects the click-through rates and advertisers’ revenue. In this paper, we answer all these questions via game theory analysis. In particular, we propose an ad dissemination protocol to protect the users’ privacy to a large extent, and formulate the interactions among all entities as a three-stage game, where each entity aims to maximize its own utility.
The main contributions of this paper are summarized as follows.
- •
We propose a privacy-aware framework for targeted advertising to motivate users, the ad broker, and advertisers to be engaged in the targeted advertising systems. This framework requires no modifications on existing targeted advertising systems, and takes the incentives of all parties into consideration. In our framework, the ad broker provides a certain amount of compensation for the users’ privacy leakage from ad clicks in order to encourage users to click their interested ads, which in turn improves the click-through rate and brings in more revenue for the ad broker and advertisers.
- •
We model the framework as a three-stage Stackelberg game, in which all entities are considered to be selfish, targeting at maximizing their own utilities by selecting optimal strategies. We analyze the cooperation and competition relationship among users, the ad broker, and advertisers, and derive the Nash Equilibrium.
- •
We further analyze the competition among advertisers who share the whole market. We model the market sharing scenario as a non-cooperative game and prove the existence of the Nash Equilibrium.
- •
We conduct numerical simulations to evaluate the proposed framework. The results verify that the utilities of all entities are notably enhanced, which provides strong motivation for the ad broker and advertisers to implement the compensation policy and users to embrace the targeted advertising.
The rest of the paper is organized as follows. Section 2 describes the system model. Section 3 introduces the compensation framework. In Section 4, we model the framework as a three-stage game and analyze the optimal strategies of advertisers, the ad broker and users. We further discuss the competition among advertisers for market sharing. Numerical results are shown in Section 5 and related works are reviewed in Section 6, followed by the conclusion in Section 7.
Section snippets
System model
In this section, we describe the system model, including the targeted advertising system and privacy sensitivity.
The compensation framework
In this section, we propose a privacy-aware compensation framework to promote targeted advertising with the consideration of privacy leakage. The target of the framework is to establish a win–win situation among all entities in the targeted advertising system. In the framework, with the promise of privacy leakage compensation, the sensitive users are incentivized to view their interested ads. The ad broker and advertisers earn more revenue from more ad clicks. It is noteworthy that our proposed
Game theory analysis
In this section, we cast the targeted advertising under the compensation framework as a three-stage Stackelberg game, and analyze the best strategies of all entities based on their utilities. We consider two advertising scenarios: independent advertisers who are engaged in independent markets and market sharing advertisers who compete against each other.
The targeted advertising game consists of three stages, as illustrated in Fig. 2. In the first stage of the game, advertisers announce the
Numerical results
In this section, we conduct extensive simulations to demonstrate the performance gain of the proposed compensation framework, and the impact of the system parameters on the performance.
Related work
Existing studies related to privacy-aware targeted advertising can be classified into three categories: targeted advertising mechanisms, privacy preservation on user’s online profiles or behaviors, and incentive mechanism designs for advertising and trading privacy.
Targeted Advertising Mechanisms. Targeted advertising mechanisms mainly focus on the strategies of advertisers or the ad broker to deliver ads to matched users. Chakrabarti et al. [22] studies the problem of displaying relevant ads
Conclusion
This paper has studied the privacy-aware targeted advertising problem, and proposed a compensation framework to encourage users to view ads of interest. The compensation framework aims to promote targeted advertising by creating a win–win situation. Under this framework, we analyze the interactions among advertisers, the ad broker, and users through a three-stage game modeling. Simulation results demonstrate the efficacy of the proposed framework, under which users are motivated to click more
Acknowledgements
The research was support in part by grants from 973 project 2013CB329006, China NSFC under Grant 61173156, RGC under the contracts CERG 622613, 16212714, HKUST6/CRF/12R, and M-HKUST609/13, as well as the grant from Huawei-HKUST joint lab.
Wei Wang is currently a Ph.D. candidate in Hong Kong University of Science and Technology. He received his B.E. degree in Electronics and Information Engineering from Huazhong University of Science and Technology, Wuhan, China, in 2010. His research interests include privacy and fault management in wireless networks.
References (33)
- et al.
Trims, a privacy-aware trust and reputation model for identity management systems
Comput. Netw.
(2010) - et al.
Privacy-preserving smart metering with multiple data consumers
Comput. Netw.
(2013) - et al.
An algorithm for privacy-preserving distributed user statistics
Comput. Netw.
(2013) - J. Yan, N. Liu, G. Wang, W. Zhang, Y. Jiang, Z. Chen, How much can behavioral targeting help online advertising?, in:...
- H. Beales, The value of behavioral targeting, in: Network Advertising Initiative,...
- Google Adwords....
- Ink tad....
- K. Purcell, J. Brenner, L. Rainie, Search engine use 2012, in: Pew Internet & American Life Report,...
- J. Turow, J. King, C.J. Hoofnatle, A. Bleakley, M. Hennessy, Americans reject tailored advertising and three activities...
- V. Toubiana, A. Narayanan, D. Boneh, H. Nissenbaum, S. Barocas, Adnostic: privacy preserving targeted advertising, in:...
Selling privacy at auction
Games Econ. Behav.
Cited by (0)
Wei Wang is currently a Ph.D. candidate in Hong Kong University of Science and Technology. He received his B.E. degree in Electronics and Information Engineering from Huazhong University of Science and Technology, Wuhan, China, in 2010. His research interests include privacy and fault management in wireless networks.
Linlin Yang is currently a M.Phil. student in Hong Kong University of Science and Technology. She received her B.E. degree of electronic engineering from Tsinghua University in 2012. Her research interests include privacy protection for advertising and network economics.
Yanjiao Chen received her B.E. degree of electronic engineering from Tsinghua University in 2010. She is currently a Ph.D. candidate in Hong Kong University of Science and Technology. Her research interests include spectrum management for Femtocell networks and network economics.
Qian Zhang ([email protected]) joined Hong Kong University of Science and Technology in September 2005 where she is a full Professor in the Department of Computer Science and Engineering. Before that, she was in Microsoft Research Asia, Beijing, from July 1999, where she was the research manager of the Wireless and Networking Group. Zhang has published about 300 refereed papers in international leading journals and key conferences in the areas of wireless/Internet multimedia networking, wireless communications and networking, wireless sensor networks, and overlay networking. She is a Fellow of IEEE for “contribution to the mobility and spectrum management of wireless networks and mobile communications”. Zhang has received MIT TR100 (MIT Technology Review) worlds top young innovator award. She also received the Best Asia Pacific (AP) Young Researcher Award elected by IEEE Communication Society in year 2004. Her current research is on cognitive and cooperative networks, dynamic spectrum access and management, as well as wireless sensor networks. Zhang received the B.S., M.S., and Ph.D. degrees from Wuhan University, China, in 1994, 1996, and 1999, respectively, all in computer science.