1.Goal The goal of this project is to use Amazon data to predict if the review is negative or positive-- If rating <=3 negative review:0, If rating >=4 positive review :1. 2.Attribute information helpful - helpfulness rating of the review [2,3], e.g. 2/3 , 2 is numerator , 3 id denominator Numerator: Number of readers who found the review is helpful Denominator: Number of readers who indicated whether they found the review helpful or not. review Text - text of the review overall - rating of the product summary - summary of the review unix Review Time - time of the review (unix time) 3. Modeling - First level modeling part - Text data : Naive Bayes Classifier, Neural Network and Logistic Regression were applies and obtained each model's predictions (train data and test data ). - First level modeling part -Non-text data : Random Forest , Neural Network and Logistic Regression were applied and generated the predictions. - Second level modeling part : All six train-data predictions and six test-data prediction were combined as new features renamed as new x_train data and new x_text data. XGboost and Neural Network were applied to predict again.