Full Citation
Title: Privacy protection algorithm based on dynamic update of data set
Citation Type: Dissertation/Thesis
Publication Year: 2013
ISBN:
ISSN:
DOI:
NSFID:
PMCID:
PMID:
Abstract: With the rapid development of various communication technologies such as the Internet, information sharing has become easier. Countries, businesses and individuals can more easily collect the useful information they need. At the same time, with the application of data mining and data publishing, the issue of privacy protection has also received more and more attention from all walks of life. In the past, most studies focused on the privacy protection methods of static data sets. However, in practical applications, the data sets that are usually released are changed with time. Therefore, how to implement privacy protection for such problems is a research focus. This paper focuses on the privacy protection of data sets with internal sensitive attribute value updates. In order to protect the privacy of data sets with this type of update, this paper introduces anonymization technology and bucket technology, and proposes on this basis. Λ-variety algorithm: Firstly, the method of reading the sensitive attribute field type of the data table is used to judge the update type of the attribute value; secondly, for different types of data set update, different bucket creation and record allocation methods are adopted. Finally, divide the equivalence class and publish it anonymously. In addition, this paper focuses on the accuracy of the data after anonymization, and proposes the (D, λ)-variety algorithm. The (D, λ)-variety algorithm adopts the greedy idea, which implements privacy protection in the data set to be distributed. It also guarantees the availability of published data sets. This paper uses the Income dataset and OCC dataset from http://ipums.org to conduct experiments. The results show that the proposed λ-variety algorithm and (D, λ)-variety algorithm can have external updates and Data sets with different categories of internal attribute value updates provide better privacy protection. In addition, the (D, λ)-variety algorithm proposed in this paper can guarantee the accuracy of the anonymized data set compared with the λ-variety algorithm proposed in this paper, but the former has a lower degree of privacy protection than the latter. Therefore, when using the privacy protection of the data set to be published, whether the λ-variety algorithm or the (D, λ)-variety algorithm is used may depend on the specific application target.
Url: http://cdmd.cnki.com.cn/Article/CDMD-10217-1014131995.htm
User Submitted?: No
Authors: Tao, Li
Institution: Harbin Engineering University
Department:
Advisor:
Degree:
Publisher Location:
Pages:
Data Collections: IPUMS USA
Topics: Methodology and Data Collection
Countries: