Mathematicians from ITMO University in Saint Petersburg, Russia, and the National University of Singapore have created an algorithm that predicts the marital status of a person with 86% accuracy. The Algorithm uses data from multiple social accounts like Twitter, Instagram and Foursquare to predict marital status. The accuracy of prediction increased by more than 17% when the data was taken from multiple accounts instead of one account.
The algorithm is capable of examining any English speaking account. Andrey Filchenkov, associate professor of Computer Technology Department at ITMO University collected and analyzed the tweets of Barack Obama and Donald trump. Based on this data the algorithm predicted accurately predicted Obama's martial status.
Also Read: Researchers analyzed Donald Trump's face and discover he is aggressive, dominant, powerful and unethical
However, it concluded Donald Trump as Bachelor. Researchers attribute this to the Trump's abnormal activity in the media. Trump does not often update his social media accounts and his assistants use twitter like Bachelor. Scientists train the algorithm to understand data input by converting the activity of users into sets or vectors of parameters such as an average tweet size, the most frequent objects in the photo, check-in distribution and so on.
Then Kseniya Buraya, a researcher at ITMO University's International laboratory "Computer technologies", process the user data with the algorithm and adapts the information to Myers-Briggs Type Indicator (MBTI). It is a scale to know the psychological types.This scale can be used to know how a person interacts with the world
Applications
According to scientists, user profiling can have a wide range of applications. It facilitates recruiters to know about the behavior of people who apply for the job
It will also help to identify illegal groups in social media accounts and also to find whether the persons prone to depression so that we can support them
No comments:
Post a Comment