Spam Detection Using TF-IDF
The Dataset used is provide by UCI, Irvine and can be downloaded from here.
The code uses TF_IDF from Scikit-learn's TfidfVectorizer() which can be imported from sklearn.feature_extraction.text. Then the extracted features are trained with Multinomial Naive Bayes. The trained model on 70-30 split gave 95%+ accuracy.