Advanced Machine Learning Model to Detect Spam on Instagram

In recent years, social media platforms like Facebook, Twitter, and Instagram have become increasingly significant as channels for the diffusion of information. One of these social media platforms that has grown in popularity recently is Instagram. Instagram is a popular network for sharing photos and videos that generate income for celebrities, businesses, and people with large followings. The site could potentially be used for illegal activities because of its enormous profit margin. One of the malicious activities on the Instagram network involves fake accounts. In this study, a hybrid approach that considers both the post's content and the images has been used for identifying Instagram fake accounts. We evaluate content spam with machine learning, and image spam with CNN models. The image dataset has been taken from Image Spam Hunter, and the model was trained using a dataset from Kaggle for categorization based on content. The proposed hybrid model has also been tested on the dataset that was scraped from Instagram. Experimental classification results show that the proposed model has a classification accuracy of 97.1%.