Optimized Deep Learning Hybrid Framework for Robust Detection of Fake News on Twitter (now X)

The rapid increase of fake news on social media platforms shows major challenges, specifically when telling real information apart from false content. This research investigates how machine learning (ML) and deep learning (DL) algorithms can be used to identify fake news. The suggested model is based on a hybrid model that combines Decision Tree (DT), Gradient Boosted Tree (GBT), and Long Short-Term Memory (LSTM) networks, utilizing ensemble methods to improve detection accuracy. The dataset for this research consists of real and fake news articles gathered from Twitter (now X), which were preprocessed using Natural Language Processing (NLP) techniques and transformed into word embedding with GloVe and Skip-Gram models. Experimental findings show that the hybrid model is significantly better than individual classifiers, achieving an accuracy rate of up to 99.7% when combining GBT and LSTM. This proposed method effectively detects misinformation, providing a strong solution to combat the spread of fake news in critical situations.