Polytechnics Create Fake Profile Detector on Facebook

The classification algorithms used in this program can determine the user’s gender and approximate age, this allows detecting false profiles in said social network.

Showing private information to strangers is increasingly common in social networks. Faced with this problem, students of the Higher School of Computing (Escom) of the National Polytechnic Institute (IPN), developed a desktop application that allows the detection of false Facebook profiles.

Through various free-code tools that allow data mining and machine learning, the young Abdul Didier Gomez Lemus and Moises Morales Almaraz managed to train a series of algorithms with which it was possible to analyze the general information of a Facebook profile, as well as as the natural language of their publications and conversations.

Morales Almaraz pointed out that with the analysis of natural language you can choose a set of independent characteristics or variables to feed the classification algorithms and these can determine the user’s sex and approximate age, later these data are compared with the information generated from the profile to identify if it is true or false.

“Although Facebook has a declaration of rights and responsibilities and a registry and account security section, there is nothing that forces users to provide real and updated data, so there is no certainty that their policies In addition, there are no penalties unless a user comes to report or report identity theft or the type of information that they share, in these cases the social network does an investigation and if they consider it they can close the account “, he assured.

Gomez Lemus indicated that a demographic study of the social network Facebook affirms that there are more 18-year-old men registered on Facebook than 18-year-old men living on planet Earth.

“This shows that there is no strict control in the creation of profiles within Facebook and probably not in other social networks, which causes distrust when someone knows a person in any of these platforms but is not sure if the information of their profile is true, or has only created this identity for cyberspace, “he explained.

The systems that are currently used for the detection of false profiles are based on the activity that accounts have over time, unlike the program created by the polytechnics that also analyzes the natural language of the person which yields better results.

The development of this application dedicated to detecting false profiles on Facebook would help all users to have greater protection against people who are engaged in doing some type of cyber crime on social networks.

 

198 Comments