lowlinkability

Lowlinkability is a property of a system or dataset in which it is difficult to associate multiple data points, events, or records with the same entity, such as an individual or account. It aims to prevent the linking of disparate observations to a single source, thereby reducing the risk of re-identification or sensitive inferences. Lowlinkability is closely related to the concepts of unlinkability and anonymity and is often pursued to protect privacy in digital environments.

In practice, lowlinkability is relevant in data publishing, analytics, online tracking, and network privacy. In data sharing, it helps ensure that records from different sources cannot be tied together to reconstruct a person’s profile. In online tracking, it mitigates the ability of advertisers or researchers to match user behavior across sites. In network privacy, techniques intended to reduce linkability include routing changes, timing obfuscation, and the use of mix networks or privacy-preserving protocols.

Common approaches to achieve lowlinkability include data minimization, pseudonymization, and anonymization, sometimes complemented by aggregation and perturbation. Privacy-enhancing technologies such as differential privacy, noise addition, and decoy traffic can also reduce linkability. In communications, protocols that separate identity from traffic or frequently change identifiers contribute to lower linkability.

Measuring linkability is an area of study in privacy research and security. Trade-offs between data utility and privacy, regulatory requirements, and potential for auxiliary information to undermine unlinkability pose ongoing challenges. Understanding lowlinkability helps inform the design of systems that balance privacy with functionality.