Detection of Fake Instagram Accounts and Spam Comments Using Machine Learning Techniques

Abstract - Social media platforms such as Instagram have become highly vulnerable to fake accounts and spam comments, which negatively impact user trust, engagement, and platform security. Manual moderation techniques are inefficient and unable to scale with the growing volume of content. This research paper presents InstaGuard, a dual machine learning–based system designed to detect fake Instagram accounts and spam comments in real time. The proposed system utilizes a Random Forest classifier for account authenticity detection based on profile metadata and a Logistic Regression model combined with TF-IDF vectorization for spam comment classification. Feature engineering techniques such as follower–following ratio, bio characteristics, username patterns, and textual n-grams are applied to enhance prediction accuracy. The system is implemented using a Flask-based backend integrated with a web dashboard that provides confidence scores and explainable outputs. Experimental evaluation demonstrates that the proposed approach effectively identifies suspicious accounts and spam comments with high reliability. The results indicate that InstaGuard offers a practical, interpretable, and scalable solution for improving safety on social media platforms. Key Words: Instagram Spam Detection, Fake Account Detection, Machine Learning, Natural Language Processing, TF-IDF, Random Forest, Logistic Regression, Social Media Security