Mobile Internet

Social media activities could help predict your next move

Social Media written with the help of scrabble chips

It is of really great concern as to how much information we should put up on social media. It is really hard to tell by the information that if it is real or not. However, the behavior on the internet might just give out everything you are trying to hide. There is a lot of difference in giving away the information by putting in the true credentials and then throwing it away through your behavior on social media. We think no one is really checking our posts or feed but there are people looking at every activity of ours and there are an algorithm and a rhythm to every single thing we post.

The fact that social media is such a vulnerable platform for people who know less about it. There are people who try and be someone else on social media and give up information about some other person who might not be on social media. This is also a case where you don’t have to be on any kind of social media to get your information on it, your friends will do it for free. There are birthday posts, some major event posts and many other things amongst them that really give away a lot of information about a friend who is not even on social media.

That said, a survey, carried out by James Bagrow, an expert in Mathematics and Statistics, who also is a professor at the University of Vermont, tells us a lot about how one account can reveal so much importance. This survey was carried out only on twitter but can easily be done on other platforms. There were many striking revelations in this survey which were absolutely stunning to look at. The number of tweets looked at this were also huge and the accounts checked and looked at were average twitter accounts.

Given all that, the data from about 13,905 Twitter accounts were analyzed. In addition to this, James and his team looked at 30.85 million tweets which is a huge number. The study was narrowed down to 927 networks which were named as ego-networks. That is one Twitter account is one network with 15 of the most mentioned accounts in that one network. This allowed them to come down to conclusions better which helped the study in a great deal.

The researchers were stunned when they could predict the future tweets of a user they did the study on. There were machine learning algorithms in use and there came out algorithms for different accounts. The algorithms could actually predict the next words you put on social media. The accuracy of prediction was up to 64%. This came down only by 3% to 61% when the predictions were made by social media activity.

There were about 95% things that could be predicted with the help of social media activity and this is one stunning revelation. This was based on twitter but the machine learning algorithms are useful for almost every social media according to Bagrow.

This is the reason why we should either actually be an open book on social media or not be there at all.

-Unmesh Phule

Comment here