GitHub - yasinda-s/TheraBot: This is a companion chatbot built using NLP. I fully developed the backend of the chatbot which consists of the CNN used to gauge the user's emotion score per each message, the ability of the bot to reply based on classified emotion, keep track of user independent scores, evaluate overall emotional status to provide with consolidation message and more!

Branches Tags

Name		Name	Last commit message	Last commit date
Latest commit History 1 Commit
Older Versions		Older Versions
.gitattributes		.gitattributes
Constants.py		Constants.py
README.txt		README.txt
TheraBotTelegramCode.py		TheraBotTelegramCode.py
TheraBotTelegramMain.py		TheraBotTelegramMain.py
responses.json		responses.json
therabot.h5		therabot.h5
tokenizer.pickle		tokenizer.pickle

Repository files navigation

In a python enviroment (3.8 or below) install the following libraries - 

json
random
os
keras==2.8.0
Keras-Preprocessing==1.1.2
tensorflow-cpu==2.8.0
pickleshare==0.7.5
python-telegram-bot==13.11
h5py==2.10.0

Once the libraries are installed, simply extract the files in the .zip file to that enviroment. 

Run TheraBotTelegramMain.py to host the chatbot on Telegram!

On the Telegram searchbar, search "unm_therabot" to message the bot.

About

This is a companion chatbot built using NLP. I fully developed the backend of the chatbot which consists of the CNN used to gauge the user's emotion score per each message, the ability of the bot to reply based on classified emotion, keep track of user independent scores, evaluate overall emotional status to provide with consolidation message an…

telegram-bot cnn tokens sequences keras-tensorflow glove-vectors tensorflow2

Readme