Troll-Detection Systems Limitations of Troll Detection Systems and AI/ML Anti-Trolling Solution
Ushma Bhatt, Divya Iyyani, Keshin Jani, Swati Mali · 2018
Trolling is a modern-day vice which has manifested itself in the burgeoning virtual world. As bullying shifts from playgrounds to social media, the need of the hour is having anti-trolling softwares in place to combat this malpractice. Many internet companies have taken steps in this direction by creating softwares or applications that can prevent trolling but none have been completely successful. The word anti-trolling has existed as a concept for a long time. It is necessary to make it a concrete reality. Most of the softwares identify abusive words and simply block them but trolls have found ways to circumvent this obstacle by using clever ways. This paper explores the currently existing “Anti-trolling systems”, their methodologies and technological challenges. In particular, we consider machine learning and existing expert systems that detect and prevent trolls. The paper concludes with a discussion of the issues faced by this technology, the current functioning and a few suggestions along with a modified architecture and the vision of a trolling-free internet.