| Title |
User De-identification Scheme for Secure Genome Data Sharing based on Local Differential Privacy |
| Authors |
엄하은(Ha-Eun Eom) ; 박영훈(Young-Hoon Park) |
| DOI |
https://doi.org/10.5573/ieie.2022.59.11.59 |
| Keywords |
LDP; Genetic data; Data privacy; Partial encryption; Lightweight |
| Abstract |
Recently, the production of genetic data has been explosively increased due to the development of genetic analysis technology and information technology, and as th use of genetic data increases in the medical field, the demand for sharing is also steadily increasing. because genetic data represents an individual’s unique characteristics, it must be managed safely and, even when shared, must be transmitted in a secure way to authorized users. however, once the genetic data is shared, the risk of exposure to genetic data increases because it is almost impossible to manage the shared data. in addition, the size of genetic data is very large, ahout 200~300G. it will take a lot of time if the existing encryption technology is applied to the genetic data. In order to solve these problems, in this paper, we propose a technique of adding noise using Local Differential Privacy(LDP) to identifiable part of a base sequence that is different from a reference genome sequence. since an irreversible noise addition method is used instead of the existing encryption method, it will be difficult for users shared genetic data with added noise to find out the original data. In addition, since noise is added to only a small portion that can identify an individual, not the entire genetic data, the execution time for applying the security technology will be drastically reduced. In other words, it will be possible to solve the problem of efficiency in managing and sharing genetic data and the problem of privacy after sharing. In the latter part of this paper, security and efficiency are verified through mathematical proofs and experiments. |