• Login
    View Item 
    •   Shocker Open Access Repository Home
    • Graduate Student Research
    • ETD: Electronic Theses and Dissertations
    • Master's Theses
    • View Item
    •   Shocker Open Access Repository Home
    • Graduate Student Research
    • ETD: Electronic Theses and Dissertations
    • Master's Theses
    • View Item
    JavaScript is disabled for your browser. Some features of this site may not work without it.

    A method for generating near optimal t-closed equivalence classes

    View/Open
    thesis (1.077Mb)
    Date
    2021-05
    Author
    Vitalapura, Spandana Siddaramanagowd
    Advisor
    Bagai, Rajiv
    Metadata
    Show full item record
    Abstract
    A huge volume of data emanates from various digital sources, this data having sensitive information is being stored and released for purposes that serve the common good through the advancement of knowledge. Organizations often need or want to publish a subset of the sensitive data they collect for regulatory or research purposes, any kind of misuse of this information creates a critical threat to ones' individuality, and to save the privacy of published data we have a research area titled as Privacy Preserving Data Publishing. Privacy Preserving Data Publishing is focused on developing methods of anonymiz- ing sensitive data such that it can be published without compromising the privacy of the individuals the data represents. One privacy guarantee that has gained recent popular- ity, t-closeness, partitions the data into equivalence classes in which the quasi-identifying attributes of the contained records are made indistinguishable from one another and the distribution of sensitive attributes within each equivalence class is guaranteed to be within a given threshold t of the distribution in the whole table. In this thesis, we present a method to achieve t-closeness for a single sensitive at- tribute, which yields us equisized equivalence classes having uniformly distributed sensitive attribute values and each of the equivalence classes satis es a lower t value. The rst step of our algorithm is forming a frequency distribution table from the input data where each sensitive attribute value is arranged in descending order of their frequency. The second step is stacking and dealing of records, here stacked records in the frequency distribution table are cyclically dealt to each equivalence classes. The third step is nding the distribution of sensitive attributes in each equivalence class and using earth movers distance to nd t value for each equivalence class. Compared to other existing methods to achieve t closeness, this method generates equisized equivalence classes having uniformly distributed sensitive attribute values which also takes care of minimal data loss and great data utility.
    Description
    Thesis (M.S.)-- Wichita State University, College of Engineering, Dept. of Electrical Engineering and Computer Science
    URI
    https://soar.wichita.edu/handle/10057/21607
    Collections
    • CE Theses and Dissertations
    • EECS Theses and Dissertations
    • Master's Theses

    Browse

    All of Shocker Open Access RepositoryCommunities & CollectionsBy Issue DateAuthorsTitlesSubjectsBy TypeThis CollectionBy Issue DateAuthorsTitlesSubjectsBy Type

    My Account

    LoginRegister

    Statistics

    Most Popular ItemsStatistics by CountryMost Popular Authors

    DSpace software copyright © 2002-2022  DuraSpace
    DSpace Express is a service operated by 
    Atmire NV