Privately Evaluating Contingency Tables with Suppression

抄録

Contingency tables are the most prevalent forms of statistical data. In this paper, we present a secure protocol for privately conducting contingency tables with suppression. We operate in a multi-party outsourcing setting where an analyst wishes to obtain the contingency table from data collected from several data contributors. The computation of con- tingency tables and suppression are outsourced to a public server. Our protocol requires one round interaction between the analyst and the cloud server, that is, the analyst sends a query to the cloud and then the cloud responses the analyst with the suppressed contingency table. We assume the server behaves semi-honestly and then the security of our protocol follows even in the presence of a malicious analyst. We implement the protocol and demonstrate that we can evaluate the contingency table with thousands of data in 8 minutes.

収録刊行物

関連プロジェクト

もっと見る

詳細情報 詳細情報について

  • CRID
    1050292572112979328
  • NII論文ID
    170000173804
  • Web Site
    http://id.nii.ac.jp/1001/00175818/
  • 本文言語コード
    en
  • 資料種別
    conference paper
  • データソース種別
    • IRDB
    • CiNii Articles
    • KAKEN

問題の指摘

ページトップへ