HYBRID CNN-DRIVEN RANDOM FOREST AND FASTTEXT EMBEDDINGS FOR UNMASKING AI-GENERATED TWEETS

This study introduces a robust framework for detecting fake tweets using a dedicated Twitter fake tweet dataset, combining natural language processing (NLP) and machine learning to improve classification accuracy. Traditional manual detection methods are limited by scalability issues, subjectivity, and the inability to effectively identify subtle linguistic or contextual signals in vast volumes of social media data. To overcome these limitations, the proposed approach employs a multi-stage pipeline. It begins with comprehensive NLP preprocessing to clean and normalize the tweet content, followed by the application of FastText embeddings to convert textual information into meaningful numerical vectors. The data is then partitioned into training and testing sets using a train-test split strategy to ensure reliable evaluation. A deep learning convolutional neural network (DLCNN) is used for sophisticated feature extraction, uncovering complex patterns within the text. These features are subsequently classified using a random forest algorithm, which determines whether tweets are real or fake. The model's performance is thoroughly evaluated using key metrics to validate its accuracy and applicability in real-world misinformation detection scenarios.