Leveraging Machine Learning algorithms for Fake Profile Detection on Instagram
R. R. Arunprakaash, R. Nathiya · 2024
In recent years, Instagram has rapidly become one of the world's largest platforms for shedding pictures and videos and running businesses online. However, despite its widespread worldwide popularity, the platform faces severe challenges due to fake profiles that spread spam and offer security threats to users. Instagram is now working on Powerful machine learning systems--advanced technology for combating fraud online--to discover and break contacts with fraudulent accounts (often infringe IP rights). Using data sets from Kaggle, which present the attributes of profiles, posting behaviours, and follower demographics, we begin an exploratory analysis and data cleanup process. Following, we test the efficacy of three classification algorithms—Random forests, linear support vector machines, and K-nearest neighbours—to construct models and reciprocate them on our scrubbed data. If the random forest classifier is the most effective, it reaches over 92.5% accuracy in recognizing false accounts. This study reflects the strategic use of machine learning techniques to reinforce safety and satisfaction across social networks. Finally, the paper discusses how time dynamics, network structure, and high-level language processing may, in turn, promote the success of these efforts to benefit users.