Abstract
Most of local differential privacy frameworks target statistics on certain privacy behaviors of users, but not behavior sequence. In this paper, we explore and propose a behavior sequence mining model that satisfies the local differential privacy requirement to settle the matter. We decompose their potential behavior sequence into multiple temporal pairs that are computed by the server to infer indirectly behavior sequence of users, shrinking the statistical sample space with adjacent temporal pairs to reduce statistical errors. The experiment takes an example, trajectories of users can be inferred by their location information, to demonstrate the effect our model achieved. It shows that the model can approximate users’ trajectories under the requirement of local differential privacy.
Highlights
The data age has spawned a variety of data collecting technologies that require different data sources, including personal privacy data
The problem that traditional differential privacy models require a trusted third party to centrally deal with individual privacy data is solved by Local Differential Privacy (LDP), and its privacy protection is theoretically more rigorous than the one offered by non-local different privacy
An algorithm model that satisfies LDP privacy is designed in the paper, decomposing users' behavior sequences into multiple adjacent temporal pairs represented by a matrix
Summary
The data age has spawned a variety of data collecting technologies that require different data sources, including personal privacy data. How to ensure that personal privacy data is not infringed when collect it has gradually become the focus of information security. The problem that traditional differential privacy models require a trusted third party to centrally deal with individual privacy data is solved by Local Differential Privacy (LDP), and its privacy protection is theoretically more rigorous than the one offered by non-local different privacy. Existing LDP models can only acquire event element frequency of collected users, but not behavior sequences. An algorithm model that satisfies LDP privacy is designed in the paper, decomposing users' behavior sequences into multiple adjacent temporal pairs represented by a matrix. The Serverside could conclude behavior sequences of users in accordance with the privacy of LDP
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have