Fabrice-64 / chatbot

Project of a chat bot in the overall framework of the developer education program

Chatbot Project

Project of a chat bot in the overall framework of the developer education program

To run this program:

• Create a virtual environment in the folder named chatbot;
• Install the required dependencies from file named requirements.txt;
• In the chatbot root folder, create a .env file to host your Google API key (see details below);
• Type flask run;
• Open your browser at url http://localhost:5000/#.

To deploy this program:

• This program is currently deployed on Heroku. Therefore some files have been added to make this possible. They are the following ones:
• nltk.txt
• .nltk_packages
• Procfile
• runtime.txt

To be noticed:

Question Parsing

• This program uses nltk to parse the questions. You may be confused by some peculiarities while installing nltk. Here is a link that could be useful, such a situation were to occur: https://www.pitt.edu/~naraehan/python3/faq.html#Q-nltk-data-path Be aware that nltk may require a specific installation process when deployed on the server.

• This program uses StanfordPOSTagger to characterize the words in a question. StanfordPOSTagger works on JAVA. Therefore, for a new deployment you may have to install a SDK Java on the remote server.

Google API key:

• The API Key is not available on GitHub;
• The user should get his own key;
• Then he creates a file named .env in chatbot folder;
• Google Key should be named: GOOGLE_API_KEY
• The link to this confidential key is set in app.controller.api_folder.config.py
• For a deployement, don't forget to add this key to the environment variables of your server.

XSS exploit:

• XSS exploit risk is tackled with by a specific script : DOMPurify, available at: https://github.com/cure53/DOMPurify