From 2013, after the revelations of Edward Snowden on mass surveillance system operated by the National Security Agency, the concept of metadata has invaded the public discussion.

Metadata refers to, literally, the information regarding the data.
Intelligence agencies claim that these metadata are basically harmless and do not reveal any sensitive information of users.
Journalists and researchers, however, have long argued the opposite.
Metadata refers to, literally, the information regarding the data.
Whatever communication, whether by phone or the Internet, leaving a trail of metadata.
In the case of telephone conversations metadata they include the phone number of the two subjects that you are calling, place and duration of the call, the IMEI code (device ID) and IMSI (user identification mobile) associated with devices.
The three researchers have developed an application specific, Metaphone, which allowed them to collect metadata from smartphone users who have decided on a voluntary basis, to participate in the study.
The total number of participants is 823, for a total of 251,788 calls and 1,234,567 analyzed text messages.
Using the information available on Google Places, Yelp and Facebook, have succeeded in identifying, using a simple automatic comparison process, 32% of the numbers called by the participants and present in their dataset.
In fact, many shops, hospitals and financial institutions provide their telephone numbers on these three platforms.
In some cases it was even possible to identify individuals because they made public on Facebook on their phone number.
Using, instead, a manual search of the individual numbers on both Google and about Intelius, a paid registry that allows to obtain public information for each individual, were able to identify up to 65% of the numbers.
The source code of the author's Messenger chat.
The data in the 'LastActiveTimes' variable are paired as follows: the number of "-" is the user ID, and the number that follows indicates the access time.
With the information gathered by this identification, the three scholars have also identified the home of about 241 130 participants who showed the city where they lived now on Facebook, required for the verification of their hypothesis, and who had at least 10 calls to companies that They had been identified with the previous method.
Localization was performed without using data from the GPS, the wireless network or from the antennas for cellular signal, thus showing how you can apply clustering algorithms, such as DBSCAN used on the basis of only the metadata of calls and SMS, in the study, to predict the location of a person.
The study, however, does not stop there and shows how metadata from phone it is possible to deduce many other sensitive information.
The three researchers could properly infer that a participant had cardiac arrhythmia.
On the basis of its metadata, the subject had received a long phone call from a cardiology unit of a medical center and had made short calls to a call-center service for a cardiac arrhythmia monitoring device.
The three researchers also claim that, regardless of whether deductions are completely accurate or not, the ability to possess highly sensitive information is a serious threat to privacy.
In fact, sensitive information that can be reconstructed from calls made to health centers, financial services, offices of political parties, to religious institutions reveal personal traits otherwise hidden, such as health problems or religious orientation and have a dramatic and dangerous impact on individual privacy.
This study shows how the phone metadata are highly interconnected, allowing easy identification of individuals by providing the ability to deduct personal relationships, places where you live and sensitive data, undermining the integrity of the user's privacy.
The metadata, however, are also linked to the activities that are carried out on the internet: IP addresses, senders and recipients of e-mail, the time at which we perform the login and session duration are some examples of available information.
This metadata can reveal tastes in shopping online or even, as shown by Sren Louv-Jansen, sleep habits.
Louv-Jansen, a programmer of Copenhagen, has developed a code that allows you to deduct the sleep patterns of their friends from the information available on Messenger, the Facebook app.
In the web version of the app shows the time that the other user made the last recently accessed.
By analyzing the source code of the page, Louv-Jansen has identified a list of identifiers of users joined what time of their last login.
By collecting this data then at regular intervals of 10 minutes it was able to reconstruct the sleep habits of his friends.
If you present an anomaly in the pattern of hours of sleep during the week, in fact, one might conclude that you have slept in someone's home or been out to dinner or to a party.
Users often log on to Facebook just before falling asleep and only riaccedono the next morning, barely awake.
This way you can reconstruct a detailed picture of the times in which you sleep: you can see which friends are to wake up early to go to work, which ones stay up late or even how to spend the weekend.
By collecting this type of data for a bow consisting of time may allow to infer sensitive information.
If you present an anomaly in the pattern of hours of sleep during the week, in fact, one might conclude that you have slept in someone's home or been out to dinner or to a party.
All sensitive information that harm the right to privacy.
Logically, as they point to the three researchers, the results obtained from them, but you can also consider those related to login on Facebook, are conservative: they are the result of analysis using a moderately sized database.
The access to the metadata of a greater number of people increases the risk of privacy breaches, facilitating the process of deduction and analysis of the information contained therein.
It does not present any problem of lack of technologies to analyze these data, as proven in the research of Stanford in the case of Louv-Jansen, who, based on simple algorithms and code lines, they were able to infer a fair amount of information.
The metadata mass gathering held by intelligence agencies, which can gather not only in large quantities but also take advantage of specific and advanced tools to perform further analysis on them, should concern us.
The idea that metadata are harmless is dangerously false.
Even General Michael Hayden, director of the NSA from 1999 to 2005, speaking at a debate at The Johns Hopkins Foreign Affairs Symposium in April 2014, has recognized their importance and said that the US government is also based on metadata to kill enemy targets.
Then continue to hold harmless the metadata, free from material and wholly unnecessary information, as it does not include the content of the conversations, it is a serious threat to privacy.

From Vice