Efficient Enforcement of Privacy Preferences for Human Digital Twins in Multi-User/Provider Scenarios

The Digital Twins (DTs) sector is expected to surpass six billion U.S. dollars by 2025, with the Human Digital Twin (HDT) being a prime example. HDTs are used in various applications, such as personalized medicine, the metaverse, and education. However, the materialization of HDTs can be costly and lead to delays in HDT-based services. To mitigate this cost, we propose a strategy, HDT-ViewMat, to identify the portions of an HDT that should be pre-materialized. The proposed strategy analyzes the process/workflow that requires HDT data to estimate the probability of its tasks being executed, also considering the compliance of the privacy policies of the task executor with users' privacy preferences. Specifically, we consider both simple scenarios, where a single user, aka a single HDT, executes the process/workflow, as well as more complex HDT usage scenarios, that is, where a group of users jointly execute the process, like attending an event or a virtual meeting in the metaverse. This latter requires considering several privacy policies and preferences in deciding which portion of each involved HDT has to be pre-materialized. At this purpose, we propose a strategy to identify a single policy, representative of the task, by solving an optimization problem that minimizes the costs for users and maximizes the gains for service providers. We conduct a set of experiments that show the feasibility of the proposed approach.