FraudLens-Fraud App Detection Using Sentimental Analysis

This project develops a Fraud App Detection system using user review analysis and Machine Learning, employing a Random Forest Classifier trained on labelled reviews categorized as fraudulent or safe. Text reviews are converted to numerical data using TF-IDF vectorization for efficient prediction. The user interface, built with Gradio, features a sleek black-themed design where users can upload app images and CSV files containing reviews for analysis. After processing, the system predicts whether the reviews indicate a fraudulent or safe app, displaying results through visual charts and providing downloadable predictions. By combining NLP and Machine Learning, this project offers a practical solution for identifying harmful apps based on user feedback, enhancing security, promoting user safety, and assisting app stores in maintaining a clean ecosystem. Key Words: Fraud App Detection, Machine Learning, Random Forest Classifier, TF-IDF Vectorization, Natural Language Processing (NLP), User Review Analysis, Gradio Interface, App Security, Fraudulent App Identification, Safe App Classification.