IRLab@IITBHU at HASOC 2019: Traditional machine learning for hate speech and offensive content identification

Abstract

In this paper, the results obtained from the Support Vec- tor Machine, XGBoost method by IRLab@IIT(BHU) on HASOC shared task-organized at FIRE-2019 are reported. The HASOC shared task has three subtasks, namely Hate speech identification, Offensive language identification and Fine-grained classification for the English, Hindi and German languages. The best result for English is obtained after apply- ing Support Vector Machine, XGBoost with a frequency-based feature for hate speech and offensive content identification. © Copyright 2019 for this paper by its authors.