Unmasking Deception: Unveiling Fake Reviews with ML
Computer Science
Project Results
Developed a sophisticated system using machine learning techniques to detect fake online reviews with 83% accuracy, aiding in consumer decision-making and promoting integrity in online platforms.
Project Description
Online reviews play a pivotal role in consumer decision-making, yet distinguishing between genuine and fake feedback remains a daunting challenge. My project endeavors to address this issue by developing a sophisticated system capable of detecting fraudulent reviews within diverse categories. I used an extensive dataset of fake-computer generated reviews and real reviews to build a classification model.
Leveraging Python libraries including Numpy, Pandas, and NLTK, I preprocessed the review data, employing techniques such as punctuation removal, lowercase transformation, stopwords elimination, and stemming. After text vectorization and normalization was done, I deployed different machine learning methods like Logistic Regression, Support Vector Classifier, and Random Forests to sort reviews into two groups: real ones made by people and fake ones created by computers. These methods helped me train my computer to spot the differences between genuine and artificial reviews. This project aims to contribute to the advancement of integrity in online platforms by showcasing the effectiveness of machine learning techniques in identifying and combating fake reviews.
Mentee
Jay Bansal
Remarks
Working on this project was super cool! With the mentor guiding us, I got to dive into machine learning and figure out how to spot fake reviews. I made algorithms that could tell real reviews from fake ones, which was awesome. It was tough at times, but the mentor was there to help me understand the code. One time, I was confused about text vectorization and normalization, but my mentor explained it in a way that made it easy to understand. She really knew her stuff and showed us the ropes, so we could explore machine learning and make our project even better. Thanks to her, we learned a lot about spotting fake stuff online and making the internet a more trustworthy place.