RACISM DETECTION IN TWEETS USING STACKED GCR-NN WITH SENTIMENT ANALYSIS

Abstract

Racism detection on social media, particularly on Twitter, has become a critical challenge due to the volume of harmful and discriminatory content. This paper presents a system using Stacked GCR-NN (Gated Convolutional Recurrent Neural Networks) combined with Sentiment Analysis to classify tweets into three categories: Not Racist, Direct Racism, and Indirect Racism. Convolutional layers capture local linguistic patterns while recurrent layers model sequential dependencies. A sentiment analysis layer further enhances detection by identifying negative emotional tones. The system achieves strong performance in distinguishing both explicit and implicit racist content.