In this project, I'm going to build a spam filter for SMS messages by using the Multinomial Naive Bayes Algorithm. My goal is to build a spam filter that classifies the new SMS messages with an accuracy greater than 80%. To train the algorithm, I'll use this dataset, created by Lyle Pratt, from data.world. The dataset is in .txt format. So I have converted it into a .csv file. The dataset contains 5,574 SMS that are already classified by humans.