Abstract:

Because of social media's dominance in the sociopolitical scene, numerous current and new types of racism emerged on the platform. Racism has appeared on social media in several forms, both hidden and open, hidden via the use of memes and open through racist statements made under false identities to provoke hate, violence, and societal instability. Racism, although frequently connected with ethnicity, is now prospering on the basis of colour, origin, language, culture, and, most crucially, religion. Social media thoughts and statements inciting racial tensions have been seen as a severe danger to social, political, and cultural stability, as well as the peace of several nations. As a result, social media, which is the major source of racist beliefs propagation, should be watched, and racist statements should be recognised and banned as soon as possible. The purpose of this project is to identify racist tweets using sentiment analysis of tweets. Because of deep learning's improved performance, a stacked ensemble deep learning model is created by merging gated recurrent units (GRU), convolutional neural networks (CNN), and recurrent neural networks RNN, which is known as Gated Convolutional Recurrent- Neural Networks (GCR-NN). In the GCR-NN model, GRU is at the top for extracting acceptable and conspicuous characteristics from raw text, while CNN extracts key aspects for RNN to produce correct predictions. Obviously, numerous tests are carried out to study and assess the performance of the proposed GCR-NN within the context of machine learning and deep learning models, demonstrating that GCR-NN has better performance with enhanced 0.98 accuracy

Description:

Racism, social media, online abuse, Twitter, deep learning.