Blood donation management system

Abstract

In today's digital landscape, content creators face numerous challenges in interpreting feedback from diverse audiences, particularly when comments are expressed in code-mixed languages such as Tamil and English, often accompanied by emojis. This research introduces a novel approach for sentiment analysis and sarcasm detection in Tamil-English code-mixed text. By integrating Bidirectional Long Short-Term Memory (BiLSTM) networks for handling code-mixed language and Convolutional Neural Networks (CNNs) for recognizing sentiment and sarcasm patterns in emojis, our approach aims to help content creators decipher the true sentiment behind multilingual comments, accurately capturing nuances conveyed through both textual and visual cues.