Purpose:<\/strong> <\/p>\n We are seeking a comprehensive dataset that includes mouse movement data for the purpose of distinguishing between human users and automated bots in web-based CAPTCHA systems. The goal is to develop and refine machine learning models that can accurately identify bot-like behavior based on mouse interaction patterns, enhancing the security and effectiveness of CAPTCHA systems.<\/p>\n Dataset Requirements:<\/p>\n Mouse Movement Data: Raw data capturing mouse coordinates, velocity, acceleration, and direction changes as users interact with a web page.<\/p>\n Click Event Data; Records of click positions, timing, and frequency to analyze the decision-making process and interaction speed.<\/p>\n Human vs. Bot Interaction: Clear distinction between data generated by human users and data generated by automated scripts (bots). This will allow for supervised learning and model training.<\/p>\n Time-Series Data: Sequential data capturing the timestamp of each mouse event to analyze the flow and pattern of movements.<\/p>\n Behavioral Biometrics: Data capturing user-specific behaviors that might indicate human-like randomness or bot-like precision in interactions.<\/p>\n Variety of Interactions: Diverse interaction scenarios, including different types of CAPTCHA challenges (e.g., image recognition, text entry) and general web browsing activities.<\/p>\n<\/div>\n submitted by \/u\/RareNeedleworker832 <\/a>
[link]<\/a><\/span> [comments]<\/a><\/span><\/p>
