18 lines
1.9 KiB
TeX
18 lines
1.9 KiB
TeX
\chapter{Conclusion and future work}
|
|
\label{ch:con}
|
|
Continuous publishing of data, also known as time series, has found over the past decades several application domains, including healthcare, smart building, and traffic monitoring.
|
|
In many cases, time series contain personal details (and are usually geotagged), and thus their processing entails privacy concerns.
|
|
%The processing/publishing of user-generated data in the form of time series, poses privacy risks to the individuals involved.\kat{the deterioration of the quality that we deal with here, is because of the privacy protection of the data, not because of the processing/publishing. I rephrased, together with the next sentence.}
|
|
Several methods have been proposed in order to protect the privacy of individuals while processing their data, but cannot avoid to deteriorate arbitrarily the quality therein. Out of these methods, we distinguish differential privacy, which quantifies the balance between user protection and data utility by a factor $\epsilon$.
|
|
|
|
In this thesis, we have concentrated on continuous user-generated data publishing. \kat{say exactly what case you covered, this is way to general. and connect it to the differential privacy that you previously mentioned, otherwise it seems irrelevant.}
|
|
We have studied the relevant literature with special emphasis on data correlation.
|
|
Furthermore, we explored ways to provide configurable protection in such settings and developed relevant solutions.
|
|
|
|
Next, we summarize this thesis in the individual chapters by describing our contribution to the problems surrounding quality and privacy in user-generated Big Data\kat{??? be specific, this is the conclusions chapter. 'The problems surrounding quality and privacy in user-generated Big Data ' means nothing.. }.
|
|
Subsequently, we discuss interesting perspectives and open questions for future investigation.
|
|
|
|
|
|
\input{conclusion/summary}
|
|
\input{conclusion/perspectives}
|