Easy-Sharing: A Personalized Privacy Diffusion Strategy Generation Method Based on Risk-Return Trade-Off

In the realm of social networks, individuals frequently engage in self-disclosure to gain social influence. However, this behavior simultaneously exposes users to significant privacy risks. Social platforms typically allow users to establish privacy diffusion control strategies, such as selecting the audience for their information. Nevertheless, the lack of privacy awareness and extensive social connections make this task time-consuming and labor-intensive for users. To address this challenge, this paper presents an automated method Easy-Sharing to assist in the effective selection of appropriate seed nodes (initial recipients), aiming to balance social influence and privacy risks according to the user’s privacy preference. This method leverages an advanced VAE-based model to predict information diffusion and proposes a privacy strategy generation method based on a risk-return trade-off. This research contributes to the development of more user-friendly and privacy-conscious social networking tools, ultimately enhancing user experience and safeguarding personal information.