This project concerns the text-to-emotion classification and fake-news detection, based on this classification. For the moment, the accuracy on test DS is order of 85%. Working on improvements (feature generation algorithms) and model-tuning (LGBM hyperparameter tuning and DL implementation) Dataset will be stored here: https://drive.google.com/drive/folders/1UqHcuW0foCNoiCJLQ-C7Ns-Mut-VQESh?usp=sharing