Twitter Spam Dataset

This dataset contains more than 78,000 tweets with 22 derived features for spam detection research. Tweets were collected using the SNScrape library, and a portion was manually annotated while the remaining were annotated using a self-training machine learning algorithm. The dataset is balanced between spam and non-spam classes, making it suitable for machine learning model training and evaluation. Potential Applications: Spam detection on Twitter Machine learning classifier benchmarking Research on social media misuse