Bot Detection Features

This dataset contains extracted features from tweet histories used for automated bot detection on social media. The features encompass semantic, behavioral, and temporal attributes, including linguistic patterns, tweet frequency, engagement metrics, and account activity trends. The dataset has been processed for use in machine learning-based classification, particularly employing Support Vector Machines (SVM) and k-means clustering for bot vs. human classification. Additionally, it supports uncertainty quantification using conformal prediction, ensuring valid and well-calibrated confidence measures. The dataset enables researchers to explore bot detection methodologies, misinformation spread analysis, and adversarial robustness in social media security. It is provided in a structured format, including labels for bot and human accounts, facilitating further experimentation and benchmarking. This resource is intended to support advancements in social media security, automated content moderation, and adversarial bot analysis.