Traditional differential privacy techniques typically centralize the original dataset in a data center and then release statistics that satisfy differential privacy. This approach is referred to as centralized differential privacy. It relies on the assumption of a trusted third-party data collector-that is, the data collector is assumed not to misuse or leak users’ sensitive information. However, in real-world scenarios, such trust is often difficult to guarantee. To address this issue, local differential privacy (LDP) has emerged. LDP redefines privacy guarantees at the user level and shifts the data perturbation process to the user’s local device, ensuring that the data received by the data collector already satisfies differential privacy. Because of its minimal trust assumptions and strong practical applicability, LDP has been widely adopted by companies such as Google, Apple, and Microsoft.