A Quest for Harmony between Privacy and Personalization: A User-Centric Approach

点击翻译按钮获取中文摘要

We live in an age in which the dependency on technological tools is inescapable. At the same time, privacy-related issues are emerging in a way that we are at the breakpoint of losing control over our own data. Indeed, too often, people jeopardize the privacy of their personal data, sharing their personal data and preferences, in order to obtain various forms of service, such as better and faster recommendations.Surveillance is everywhere! We switched from “Big brother is watching you” to “Big mother is watching you”. Our society is being mothered; thanks to its benevolent mother the world has become a state of docility, voluntary servitude, and transparency leading to irreversible renouncement of privacy.As stated by Tim Berners-Lee, “We lose out on the benefits we could realise if we had direct control over this data and chose when and with whom to share it”. Today, more than ever, users may need to keep control over their personal data. In this research program, I propose to address the balance between privacy and personalization inherent to e-learning and e-commerce, in a platform that includes a social media environment. My main objective is to design and build a user-centric holistic platform that can preserve an individual's privacy online. Instead of using one privacy protection system for each online service he may use, I propose one holistic tool that can help him protect his digital life. To do so, the tool will enable the user to keep his data locally and disclose it only when needed, and furthermore, only what is needed.Therefore, I shall develop a holistic system architecture and protocol for building hybrid recommender systems that simultaneously employ demographic, collaborative and content-based filtering, all the while maintaining privacy objectives. I shall use the division of trust principle along with cryptographic tools as well as differential privacy to reach a balance between privacy and personalization. I propose to develop a multi-agent platform based on intelligent agents that help social media users manage their personal data.The social goal of the platform is to learn the privacy behaviour of the user in order to give him personalised recommendations. Ultimately, not only the platform will increase the user's security awareness and attitudes towards privacy, but it will learn from his profile and those of his peers for education and ethics purposes.Furthermore, the multi-agent platform is “altruistic”: it is inspired by a co-utility-based approach, in which each user contributes in preserving the privacy of others, and eventually helps in making the platform even more intelligent and customizable.This program will culminate with the realization of a proof-of-concept implementation in three contexts that I have studied for more than 20 years, namely social media, e-learning and e-commerce.