Structure of the project

Steps 1 and 2 can be skipped if, for the process, is used the file dataframe.csv or is followed the step 3

1. Use the scraper selen.py to download debates from Kialo, that are saved as csv file
2. In the notebook argumentative-debates.ipynb the debates (in format csv with ';' as separator) are loaded and "cleaned" (according to what is written in the Proia-Spurio-Urbinati_NLP-Exam.pdf file)
3. The dataset can be build and cleaned as in point 1 and 2 or can be loaded adding as shortcut this Google Drive directory in the path drive/MyDrive/NLP/project (or load into Colab only the dataframe.csv and change the load path)
4. Models
   • Neural Network models, based on Bi-LSTM
   • Machine Learning models: Linear Regressor, Decision Tree Regressor, Random Forest Regressor, Support Vector Regressor