This repo contains our Bayesian Methods class project. It consists of building a fake review detector by using a Bayesian approach that not only outputs predictions, it also provides a meaningful way of analizing the predictors used to make such predictions.
Although we started working on all the dataset we were given by Dr. Rayana, we realized that the behavioural features used for modeling didn't make sense when the reviewer only has written one review. Therefore, all the interesting work is started working.
During this project, we made some interesting findings, which you can read about in our report and presentation.