Distinguishing Hate Speech from Sarcasm
Published in 2022 International Conference for Advancement in Technology (ICONAT), 2022
In the complex landscape of online communication, distinguishing between sarcasm and hate speech is critically important. Misinterpretations can lead to significant social consequences. In this paper, we introduce a novel dataset and employ advanced machine learning models, including BERT, to better classify textual content across various contexts, achieving an accuracy of around 80%. Our approach helps in accurately interpreting user intent, contributing significantly to the field of sentiment analysis. This dataset comprises four categories—Hate Speech, Offensive, Sarcasm, Normal—evaluated using various state-of-the-art models. The highest accuracy was achieved using BERT.
Recommended citation: Srinivas Pavan, et al. (2022) "Distinguishing Hate Speech from Sarcasm" in Proceedings of the 2022 International Conference for Advancement in Technology (ICONAT). https://ieeexplore.ieee.org/document/9726093