The History and Mathematics of Codes and Code Breaking

Tag: datamining

Should We Mine Students’ Data?

In his essay entitled “Mining Student Data Could Save Lives”, Michael Morris argues that schools should surveil the online activities of their students in order to  predict and prevent acts of violence before they occur. Because all network traffic goes through the school’s systems, the IT department in a school can monitor the online behavior of any student, which certainly has the potential to help administrators watch for behavior indicative of intent to act, as well as identify at-risk individuals who managed to fly under the radar before. Morris argues that despite the intrusion of privacy, the monitoring system should be implemented because of its potential to prevent crime. Furthermore, he justifies the lack of privacy by arguing that individuals willingly give up personal information on social media and that companies already collect data for marketing purposes.

Although I do agree with the premise of doing whatever possible to save lives, I don’t agree with the blanket surveillance that Morris seeming to be advocating for. Ignoring the violations of privacy, I don’t believe that looking for patterns in online behavior is an efficient method of determining risk. There are bound to be many false positives from students who don’t actually pose any risk that would waste the school’s resources while they investigate, or worse, students who have malicious intent but don’t exhibit any observable patterns online would completely fly under the radar. In a similar situation, the Patriot Act was passed after 9/11 which allowed public surveillance with the goal to prevent acts of terrorism. However, out of all of the criminal referrals that resulted from investigations, none of them were for terrorism. Acts of terror obviously still occur in the country despite the FBI’s resources, which leads me to question the effectiveness if implemented by a school system. I believe that a more effective method would be targeted investigations of online behavior based on tips from people, which would allow schools to focus their resources in a more efficient manner.

Mining Student Data – To protect students or invade students’ life

In his essay, Michael Morris states that through mining student data, threat-assessment teams of universities have a great chance and plausible accuracy to predict possible violent behaviors with mining algorithms. As a result, it is responsible for every university to monitor students’ academic record and prevent every possible tragedy.

Undoubtedly, in no way are we supposed to turn blind eyes to the fact that as the development of statistics and computer science, the mining algorithms is remarkable in this Internet era. Like the way Amazon know what productions we are interested in and may purchase and recommend them in our app,  threat-assessment teams are probably detected most of possible violent behaviors before they come true. So mining student data can be a effective way to prevent those terrible issues like suicide or criminal behaviors. 

However, the accuracy of the mining algorithms is not 100 percent, or even far less than 100 percent. As Micheal Morris said in his essay, I have had my credit-card transaction declined for many times since I created in China but lived in America now. The bank monitored my transaction record and defined it as an unusual pattern of spending America dollars with my credit card. My life was heavily infected during the time my card was freezing. The protection that bank provided actually based on the inaccurate result of the mining algorithm and it took the unnecessary action. The same as my bank, Amazon usually makes inaccurate prediction and recommendation as well, that’s why our app often shows productions unrelated to our interests.

Similarly, the mining algorithms can lead threat-assessment teams to wrong direction and judge some nonviolent academic record as possible threat to campus safety. If a university take action according to that prediction and ask an innocent student to have a conversation, that will definitely affect the student’s daily life. The feeling of being monitored and offended will come to the student and prevent them from concentrating on their career. In that case, the university just invade the student’s life let alone by a way that even though the university does not take any action, it has already pried the student’s privacy.

Under the situation that the algorithm can not reach 100 percent accuracy, universities which use the data-mining technology may invaded normal students’ daily life. As a result, I disagree with the statement of Micheal Morris and consider that it is not time for universities to abuse the data-mining technology.



Mining data could cause more harm than good

Michael Morris has written an article arguing that ‘Mining student data could save lives’ for The Chronicle. Morris thinks that places of higher education should use the data they gather about their students from their servers to spot certain behavioral patterns or “warning signs” that could lead to certain situations such as terrorist attacks. He argues that the number of people killed will decrease and that this is a justification for making data public over keeping it private. I disagree with the author’s thesis because I believe that even though “data mining could save lives” it will actually cause more problems than it solves.

Once we start giving up our right to privacy of information we begin to lose track of where we draw the line between whether something should be kept private or made public. For example, in the San Bernardino shootings where Apple refused to allow the government access to the shooter’s phone. Had Apple conceded to the government’s wishes then not only does it undermine our basic human rights as citizens to privacy but also it gives the impression that any organization can gain access to any information whenever they want it in the name of security. Once we start exchanging our freedom (of privacy) for safety then these organizations, in this case universities, then this can lead to universities and other organizations requesting and compiling more data from us which just makes the term “privacy” obsolete. This huge compilation of data may not only be available to the organizations themselves but to other people with malicious intent too.

If every student agreed to let their data be used by the university or college that only creates another problem which is making sure that all that information is kept safe and secure. If a university collects data from students and this information isn’t protected well enough, thousands of people’s names, financial information, phone numbers, and other things will be available for anyone to get access to. This happened recently at Michigan State University which then lead to the administrative staff paying a ransom of $15,000 for the hackers to stop. This attack, although small, clearly shows how mining student data can make more people susceptible to crimes than the amount of lives that it could potentially save.

To conclude, while I agree that Morris’ argument that data mining could save lives, I do think the implications of mining such data not only puts more people at risk to a different variety of crimes but also, creates a gray area of what information we can actually keep private, if there is any.

Powered by WordPress & Theme by Anders Norén