Klasifikasi Emosi Terhadap Konflik Israel-Palestina Menggunakan Algoritma Gated Recurrent Unit

Abstract

The Israel-Palestine conflict intensified following the October 7, 2023, attack by Hamas on Israel, triggering various emotional reactions on social media. Emotion classification is crucial for understanding public sentiment related to this conflict. This study utilizes 9,917 tweets from platform X (Twitter) to classify emotions such as joy, sadness, anger, fear, disgust, and surprise. The deep learning algorithm used is Gated Recurrent Unit (GRU), developed with three different training and testing data splits: 70:30, 80:20, and 90:10. For text representation, Global Vector (GloVe) word embedding is employed. Given the imbalanced dataset, this study applies the Synthetic Minority Over-sampling Technique (SMOTE) to address class imbalance. The research results indicate that the GRU model with a 90:10 data split without using SMOTE achieves the highest accuracy of 75%, followed by the models with 70:30 and 80:20 splits, which each have an accuracy of 73%.