From 9dd5d13cb482411cef58dce10d04eed7a57a532c Mon Sep 17 00:00:00 2001 From: Manos Katsomallos Date: Sun, 24 Oct 2021 14:06:43 +0200 Subject: [PATCH] related: Done --- text/related/micro.tex | 8 ++++++-- text/related/statistical.tex | 7 ++++++- text/related/summary.tex | 5 +++-- 3 files changed, 15 insertions(+), 5 deletions(-) diff --git a/text/related/micro.tex b/text/related/micro.tex index b411576..18acf16 100644 --- a/text/related/micro.tex +++ b/text/related/micro.tex @@ -458,6 +458,10 @@ The goal is to minimize the information throughput and always answer users' requ They model the dependence between requests using a Markov chain, which is publicly known, where each state represents an available service. Setting privacy to ON, the user obfuscates their original query by randomly sending requests to (and receiving answers from) a subset of all of the available services. Although this randomization step makes the original query indistinguishable while making sure that the users always get the information that they need, there is no clear quantification of the privacy guarantee that the scheme offers over time. -\bigskip -\kat{Add here the comparison/contrast paragraph of microdata techniques shown previously, and your work} +% \kat{Add here the comparison/contrast paragraph of microdata techniques shown previously, and your work} +\bigskip +Our work is directly applicable to microdata, and thus it applies to most of the scenarios that we discussed in this section. +Most microdata methods in continuous data publishing rely on $k$-anonymity and its derivatives, and therefore their main point of failure is the linkage and background knowledge related attacks. +Since we base our privacy notion on differential privacy, we can efficiently tackle this challenge. +Finally, quite a few of the reviewed article consider data dependence and particularly temporal correlation, which is inherent in continuous data publishing. diff --git a/text/related/statistical.tex b/text/related/statistical.tex index 864474f..6bde13f 100644 --- a/text/related/statistical.tex +++ b/text/related/statistical.tex @@ -430,4 +430,9 @@ Whereas, increasing the discount coefficient resembles the behavior of event-lev Selecting a suitable value for the privacy budget and the discount parameter allows for bounding the overall privacy loss in an infinite observation scenario. However, the assumption that all users discount previous data releases limits the applicability of the the current scheme in real-world scenarios for statistical data. -\kat{Add here a paragraph that contrasts/compares your work with the works presented for statistical data.} +% \kat{Add here a paragraph that contrasts/compares your work with the works presented for statistical data.} +\bigskip +Most of the proposed methods in this section utilize differential privacy, on which we base our work. +However, few of them account for data dependence and particularly temporal correlation, which is inherent in time series. +In this thesis, we generally investigate the presence of correlation in data and we propose a method that accounts for temporal correlation throughout finite time series. +Last but not least, although the use case of our work focuses on microdata, it can adapt to scenarios that requite data aggregation, and thus extend its applicability. diff --git a/text/related/summary.tex b/text/related/summary.tex index 77712f9..8a89f73 100644 --- a/text/related/summary.tex +++ b/text/related/summary.tex @@ -1,7 +1,8 @@ \section{Summary} \label{sec:sum-rel} -In this chapter, we surveyed the literature around the domain of privacy-preserving continuous data publishing. -We offer a guide that would allow its users to choose the proper algorithm(s) for their specific use case. +In this chapter, we surveyed the literature around the domain of privacy-preserving continuous data publishing in microdata and statistical data. +We further categorized the works in terms the span of the data observation in finite and infinite. +Moreover, we summarize the methods for each data category in tabular form (with detailed attributes) aiming to offer a guide that would allow its users to choose the proper algorithm(s) for their specific use case. Such a documentation becomes very useful nowadays, due to the abundance of continuously user-generated data sets that could be analyzed and/or published in a privacy-preserving way, and the quick progress made in this research field. % \kat{? Don't forget to mention here the publication that you have.} % \mk{Done in the beginning}