Privacy-preserving and fraud-resistant targeted advertising for mobile devices

Mamais, Stylianos 2019. Privacy-preserving and fraud-resistant targeted advertising for mobile devices. PhD Thesis, Cardiff University. Item availability restricted.

Preview	PDF - Accepted Post-Print Version Download (4MB) | Preview
	PDF (Cardiff University Electronic Publication Form) - Supplemental Material Restricted to Repository staff only Download (2MB)

Abstract

Online Behavioural Advertising (OBA) enables Ad-Networks to capitalize on the popularity of digital Publishers in order to target users with contextaware promotional materials from Advertisers. OBA has been shown to be very effective at engaging consumers but at the same time presents severe privacy and security threats for both users and Advertisers. Users view OBA as intrusive and are therefore reluctant to share their private data with Ad-Networks. In many cases this results in the adoption of anti-tracking tools and ad-blockers which reduces the system's performance. Advertisers on their part are susceptible to financial fraud due to Ad-Reports that do not correspond to real consumer activity. Consequently, user privacy is further violated as Ad-Networks are provoked into collecting even more data in order to detect fictitious Ad-Reports. Researchers have mostly approached user privacy and fraud prevention as separate issues while ignoring how potential solutions to address one problem will effect the other. As a result, previously proposed privacy-preserving advertising systems are susceptible to fraud or fail to offer fine-grain targeting which makes them undesirable by Advertisers while systems that focus on fraud prevention, require the collection of private data which renders them as a threat for users. The aim of our research is to offer a comprehensive solution which addresses both problems without resulting in a conflict of interest between Advertisers and users. Our work specifically focuses on the preservation of privacy for mobile device users who represent the majority of consumers that are targeted by OBA. To accomplish the set goal, we contribute ADS+R (Advert Distribution System with Reporting) which is an innovative advertising system that supports the delivery of personalized adverts as well as the submission of verifiable Ad-Reports on mobile devices while still maintaining user privacy. Our approach adopts a decentralized architecture which connects mobile users and Advertisers over a hybrid opportunistic network without the need for an Ad-Network to operate as administrative authority. User privacy is preserved through the use of peer-to-peer connections (serving as proxy connections), Anonymous- download technologies and cryptography, while Advertiser fraud is prevented by means of a novel mechanism which we termed Behavioural Verification. Behavioural Verification combines client-side processing with a blockchaininspired construction which enables Advertisers to certify the integrity of Ad-Reports without exposing the identity of the submitting mobile users. In comparison to previously proposed systems, ADS+R provides both (1) user privacy and (2) advert fraud prevention while allowing for (3) a tunable trade-off between resource consumption and security, and (4) the statistical analysis and data mining of consumer behaviours.

Item Type:	Thesis (PhD)
Date Type:	Completion
Status:	Unpublished
Schools:	Computer Science & Informatics
Date of First Compliant Deposit:	7 October 2019
Last Modified:	23 Jul 2020 02:15
URI:	https://orca.cardiff.ac.uk/id/eprint/125897

Actions (repository staff only)

Edit Item