Scientists come up with algorithm to fend internet trolls




Scientists at Stanford University have developed a tool that people who make online provocative, insulting or threatening comments, can recognize. The algorithm would ‘Internet trolls’ recognize after five to ten postings.

The researchers studied three new online communities: those of CNN, the political news site and tech site IGN. They used before a dataset with 10,000 users that have ever banned because of antisocial behavior. In addition, they used a set of all published posts from these users.

The purpose of the research was to determine if the users were in their antisocial behavior. In addition, they wanted to know if the behavior was worse and to what extent antisocial users based on their behavior could be identified.

The researchers used machine learning to see how they could recognize the internet trolls. This means that they formulas applied to the datasets and each case examined how likely it is that it was a sociopathic user. By the computer to analyze these cases and thus make “train” the software can later classifiers indicate a certain probability if someone is a troll.

And that is exactly what the scientists were doing. They compared the messages that users, who were later blocked, had previously placed with those of users that were never kept. There were clear differences. Thus messages from internet trolls less legible, an analysis by means of the Automated Readability Index . In addition, spend more time trolling specific threads than regular users.

The classifier developed by the researchers, only the first five messages from a user with eighty percent probability indicate how someone later gets a spell, so state in the paper. The algorithm performs better when it has analyzed a user of the first ten messages.

This research can be written programs in the future that can automatically keep internet trolls websites, what the work of moderators can drastically reduce. However, it must therefor and for exposing details of trolling the software to be developed further. Currently, the algorithm picks someone more extreme over the line has gone namely rather than someone who has something say once placed which can not be tolerated. By analyzing more blocked users and their properties, the software can determine precisely when another is a troll.

Twitter is one of the sites that faced increasingly with trolls. The company is trying to avoid using banned phone users who intimidate other users, creating ever new accounts. For now there are no plans to suggest that Twitter users intimidating wants to ban automated.


In: A Technology & Gadgets Asked By: [19102 Red Star Level]

Answer this Question

You must be Logged In to post an Answer.

Not a member yet? Sign Up Now »

Star Points Scale

Earn points for Asking and Answering Questions!

Grey Sta Levelr [1 - 25 Grey Star Level]
Green Star Level [26 - 50 Green Star Level]
Blue Star Level [51 - 500 Blue Star Level]
Orange Star Level [501 - 5000 Orange Star Level]
Red Star Level [5001 - 25000 Red Star Level]
Black Star Level [25001+ Black Star Level]