A team of students led by professor and dean (R&D) of International Institute of Information Technology, Hyderabad (IIIT-H) has created a detector for Twitter which will spot hate speech shared on this widely used microblogging site. This is one of its kinds systems to scan Twitter for hate speech. 

This team comprises professor and dean (R&D) Vasudeva Varma, students Shashank Gupta and Pinkesh Badjatiya and also adjunct faculty Manish Gupta. The team has been working on the project titled 'deep learning for hate speech' for almost a year, reported the Times of India. 

The system that this has developed is capable of detecting abusive language, racist and sexist speech and then flag the offensive content. This system will help to filter content and to understand public sentiments to identify problems; the daily quoted the team as saying.

Earlier this month, this project was presented at International World Wide Web 2017 in Perth, Australia. The poster of ‘Deep Learning for Hate Speech Detection in Twitter’ was judged as the best poster presentation by popular votes. There were 166 submissions from across the globe, reported The New Indian Express. 

The system uses ‘supervised learning’ a widely used approach in machine learning to detect the hate speech. Various examples of different forms of hate or abuse via text are fed to the computer algorithm to identify the objectionable content.