Barker, KenMoniruzzaman, Md.2022-09-272022-09-272022-09Moniruzzaman, M. (2022). Redeem with Privacy (RwP): privacy protection framework for geo-social commerce (Doctoral thesis, University of Calgary, Calgary, Canada). Retrieved from https://prism.ucalgary.ca.http://hdl.handle.net/1880/115306Geo-social networks (GSN) are online social networks where people interact based on their location and relationship. These applications have gained popularity due to their innovative features. However, there are numerous privacy risks of using GSNs. Users may expose their mobility history to unknown third parties since many of these applications rely on collecting and sharing users' information. Business organizations encourage people to do a check-in to their store on GSNs by offering promotions and deals. Check-in is a virtual form of visiting a location. When a user performs a check-in to a business organization, the record is shared with the merchant. GSNs lack transparency in explaining how the third parties handle users' information. In practice, a dishonest merchant may use check-in histories to track the user's location. It may cause privacy breaches like robbery, disclosure of meetings, stalking, etc. In my Ph.D. thesis, I investigate privacy issues arising from deal redemption in GSNs. I perform an exploratory study on several GSN datasets to understand when people visit different types of locations. The study shows that there is a high degree of regularity in the user's check-in behavior. Since a typical deal requires multiple check-ins from the user within a short period, the user may become vulnerable to location tracking by redeeming deals. One potential solution is to minimize the volume of check-in information released when the user redeems deals. In this thesis, I propose a policy to identify redundant information that is not essential for a merchant to know and suppress them. I also explore the possibility that a merchant may apply inference attacks to recover the deleted information. Several inference methodologies have been investigated in my thesis, showing that a merchant can recover the data with high accuracy. I study an adversarial technique to improve a user's privacy by increasing the merchant's inference error. A recommendation algorithm is proposed to rank check-ins that a user can follow to redeem deals. Ranking applies various factors that people consider when choosing a check-in date, such as daily routine, the promotional value, and privacy. Results show how different user preferences map to various levels of inference accuracy. It would provide helpful feedback to users on how to change their preferences to enhance their privacy.engUniversity of Calgary graduate students retain copyright ownership and moral rights for their thesis. You may use this material in any way that is permitted by the Copyright Act or through licensing that has been assigned to the document. For uses that are not allowable under copyright legislation or licensing, you are required to seek permission.social networkprivacyartificial intelligenceArtificial IntelligenceComputer Sciencedoctoral thesis