ElizaBot_NLP_with_Spacy

The project for the NLP course: an Eliza-like chatbot created combining classic generic regex matching and specific spacy matching.

Library dependencies

• pip install python-telegram-bot
• pip install clean-text
• pip install setuptools wheel
• pip install spacy
• python -m spacy download en_core_web_lg
• pip install numpy
• pip install nltk
• pip install transformers

Files

• main.py contains the connection to the telegram API.
   Runs this file to start up the telegram bot.


• logic.py contains the main logic of the chatbot.
   Runs this file to use the terminal version of the chatbot.
    The hard_coded_case flag let the bot respond to an hard-coded input. Disable it if you want to chat with the bot via terminal.


• sentence_analysis.py contains all the functions used to analyze the user input text with Spacy.


• emotion_analyzer_classifier.py allows to load a pre-trained emotion classifier using the Transforms API.
• NLTK_Chat_modified.py is a modified version of NLTK Chat that allows me to create more sophisticated pairs/rules pattern->responses.
• responses.py contains
  • the specific patterns/responses rules ;
  • the generic patterns/responses rules (these responses are emotion-sensitive) ;
  • the memory patterns/responses rules (it allows to create a sentence given a memory frag);
  • the reflections used to reflect pronouns and eventually the verb associated to them.


• my_secrets.py should contain the bot's token given by the @BotFather.