Detecting Fake Reviews with Isolation Forest

In this project, I use isolation forest to detect fake reviews. We have a dataset with reviews from a website that sells coffee tables. I explore and preprocess the data using various statistical tools. Based on my exploration, I decide that the best approach to detect fake reviews is to develop a hybrid system. This system would combine a rule-based algorithm, an isolation forest model, and a K-means clustering model. I train two isolation forest models and evaluate and interpret their results.