LEO - Chatbot with Web Interface

Overview

LEO is a chatbot with a web interface, using natural language processing (NLP) and Facebook's Blenderbot for conversational interactions. The chatbot is designed to provide responses to user inputs based on predefined intents.

Features

• Natural Language Processing: The chatbot employs NLP techniques to understand and respond to user inputs.
• Machine Learning: The underlying model is trained using TensorFlow and TFLearn for improved conversation handling as well as a pre-built Facebook chatbot.
• Web Interface: Users can interact with the chatbot through a simple web page, enhancing the user experience.

Getting Started

Prerequisites

• Python 3.X
• Required Python libraries: os, sys, json, nltk, random, tflearn, numpy, socketserver, tensorflow, transformers

Installation

1. Clone the repository:

   git clone https://github.com/shiventi/Personal-Projects.git
   cd Personal-Projects/LEO
   

Usage:

python facebook_bot.py (make sure to pip install all libraries)

Customization:

• Customize chatbot responses by modifying the intents.json file.
• Train the chatbot with additional data by updating the training data in the script or through external files.

Project Structure

LEO/

• facebook_bot.py # Main script containing the chatbot logic and web server setup.
• intents.json # JSON file containing training data for the chatbot.
• chatbot_dependencies.py # Module for additional chatbot dependencies.
• webchat.html # HTML file for the web interface.
• favicon1.png # Image for HTML (favicon).
• leo_inaction.png # Image of LEO in action.
• model.tflearn # Saved model file (will be created once the code runs).