Tweet Prediction for Social Media using Machine Learning

Abstract

Tweet prediction plays a crucial role in sentiment analysis, trend forecasting, and user behavior analysis on social media platforms such as X (Twitter). This study delves into optimizing Machine Learning (ML) models for precise tweet prediction by capturing intricate dependencies and contextual nuances within tweets. Four prominent ML models, i.e. Logistic Regression (LR), XGBoost, Random Forest (RF), and Support Vector Machine (SVM) were utilized for disaster-related tweet prediction. Our models adeptly discern semantic meanings, sentiment, and pertinent context from tweets, ensuring robust predictive outcomes. The SVM model showed significantly higher performance with 82% accuracy and an F1 score of 81%, whereas LR, XGBoost, and RF achieved 79% accuracy with average F1-scores of 78%.

Keywords

• tweet prediction
• emotion analysis
• machine learning
• hyperparameter tuning