Feature Extraction Method Comparison in Detecting Spam Tweets Using Machine Learning
Franscesco William Gazali, Mohamad Soleh Hidayat, Ghinaa Zain Nabiilah, Lili Ayu Wulandhari · 2025
Twitter is one of the most used social media platforms, with millions of active users worldwide, faces a growing problem with spam. Spam encompasses a wide range of harmful content, including misleading information, irrelevant messages, malicious links, and deceptive schemes designed to exploit users. This type of content not only disrupts the user experience but also poses significant security risks, such as the potential spread of malware, financial fraud, and the loss of trust within the platform. Addressing this issue effectively is crucial for maintaining the platform's integrity and user safety. This research explores the application of machine learning techniques to improve spam detection on Twitter. This research leverages machine learning techniques to address this issue, utilizing features such as tweet text, retweet count, and user profiles to classify tweets as spam or not spam. By employing algorithms like Random Forest, SVM, and Neural Networks, the proposed model aims to enhance accuracy and efficiency in spam detection.