70 lines
3.0 KiB
TeX
70 lines
3.0 KiB
TeX
|
|
|
|
\section{Results\label{sec:resul}}
|
|
|
|
%\includegraphics[width=\linewidth]{figures/pdf/01_3rdparty.pdf}
|
|
|
|
This section introduces all the findings in this work. The main priority is the variable \texttt{tracking\_hints}, as the work tries to identify relationships between trackings and cookies. Every graph in this work is filtered for \texttt{tracking\_hints=yes}. Meaning the tables retrieved are larger than those visualized in this work.
|
|
|
|
\begin{figure}[h]
|
|
\centering
|
|
\includegraphics[
|
|
width=\linewidth,
|
|
]{figures/png/02_browser.png}
|
|
\caption{Tracking hints on each chromium and firefox}
|
|
\label{fig:resul:02_browser}
|
|
\end{figure}
|
|
|
|
|
|
The first figure, \autoref{fig:resul:02_browser}, illustrates the distribution of tracking hints across browsers and search engines. DuckDuckgo appears to be the Search~Engine that caches and identify most hints to tracking. Bing and Brave appear to conservative addressing any tracking hints. While only on DuckDuckGo and Google, the choice of web-browser seem to play a crucial role. Chromium addresses more hints to tracking in Google than Firefox does.
|
|
|
|
|
|
Furthermore, request cookies counts and response cookies counts will be presented in the Results below, which will be the main focus of the following results.
|
|
|
|
\begin{figure}[h]
|
|
\centering
|
|
\includegraphics[
|
|
width=\linewidth,
|
|
]{figures/png/03_proxy-request.png}
|
|
\caption{Proxy mode configuration for request cookies}
|
|
\label{fig:resul:03_proxy_request}
|
|
\end{figure}
|
|
|
|
|
|
Figures~\ref{fig:resul:03_proxy_request} and~\ref{fig:resul:04_proxy_response} illustrate the cookie counts for each tracking hint. DuckDuckGo and Brave show no cookies across all \texttt{tracking\_hints}. In contrast, only Bing and Google use cookies on entries identified as tracking hints. The number above the name of the Search~Engine indicates the number of cookies, and the left bar indicates how many instances of those are associated with a tracking hint. Both Bing and Google shows cookies count on requests, but only Google shows cookies count on response.
|
|
|
|
|
|
Figures~\ref{fig:resul:03_proxy_request} and~\ref{fig:resul:04_proxy_response} express how using a proxy influences tracking hits on the web. Using Tor as a proxy returns no cookie responses on tracking hints, while some cookies still leak through in request cookies associated with tracking hints.
|
|
|
|
|
|
|
|
|
|
\begin{figure}[H]
|
|
\centering
|
|
\includegraphics[
|
|
width=\linewidth,
|
|
]{figures/png/04_proxy-response.png}
|
|
\caption{Proxy mode configuration for response cookies}
|
|
\label{fig:resul:04_proxy_response}
|
|
\end{figure}
|
|
|
|
|
|
|
|
|
|
In \autoref{sec:appen:figures} additional visualization on cookies comparison on request and responses are added.
|
|
|
|
% \subsection{Dataset overview}
|
|
|
|
% \subsection{Exploratory Data Analysis}
|
|
|
|
% \subsection{Cookie activity by search engine}
|
|
|
|
% \subsection{Third-party request analysis}
|
|
|
|
% \subsection{Tracking-related domains}
|
|
|
|
% \subsection{Temporal patterns}
|
|
|
|
% \subsection{Outliers and anomalies}
|
|
|
|
% \subsection{Summary of findings} |