metadata

license: apache-2.0
title: 🧠AI Chatbot🤖
sdk: streamlit
emoji: 💻
colorFrom: pink
colorTo: purple
thumbnail: >-
  https://cdn-uploads.huggingface.co/production/uploads/6686260107019f3fe482ce08/xfpa6MidZ5aE9OEP96pi5.jpeg
short_description: The System on Real-Time Intent Recognition and Conversations
sdk_version: 1.44.1

🤖 AI-Powered Chatbot using NLP

📌 Introduction

This project is an AI-driven chatbot, developed as part of my AICTE-Shell Internship. The chatbot leverages Natural Language Processing (NLP) and Deep Learning techniques using BERT to provide intelligent responses based on user queries. The chatbot is trained on an Intent JSON dataset and fine-tuned to enhance accuracy.

🔗 Deployed Application: 🧠AI Chatbot🤖

🎯 Project Goals

✅ Implement AI & NLP techniques for intelligent conversation.
✅ Explore BERT-based Deep Learning for chatbot development.
✅ Develop a context-aware chatbot with high accuracy.
✅ Enhance text preprocessing, model training, and deployment skills.
✅ Deploy an interactive chatbot web app using Streamlit.

📂 Dataset Used

The chatbot is trained on a custom Intent JSON dataset, which includes:

User Queries & Responses: Predefined conversations.
Intent Classification Data: Labeled conversations for accurate intent detection.
Pretrained BERT Model: Fine-tuned for improved understanding.

📊 Methodology

Step 1: Data Collection & Preprocessing

🔹 Loaded and cleaned Intent JSON dataset.
🔹 Tokenized text data using BERT tokenizer.
🔹 Converted labels to categorical format for training.

Step 2: Model Selection & Training

🔹 Used BERT (Bidirectional Encoder Representations from Transformers).
🔹 Implemented deep learning-based intent classification.
🔹 Trained on multiple epochs & tuned hyperparameters for optimal accuracy.
🔹 Evaluated training & validation accuracy to ensure model performance.

Step 3: Chatbot Development & Integration

🔹 Built an Intent Recognition Model using BERT for Sequence Classification.
🔹 Designed a Response Generation Mechanism for accurate replies.
🔹 Integrated trained model into a Streamlit & HuggingFace web app for user interaction.

Step 4: Deployment & User Interaction

🔹 Saved and exported the trained BERT model for real-time inference.
🔹 Deployed chatbot as a Streamlit as well as HuggingFace web app.
🔹 Implemented real-time conversations with NLP-powered responses.

🔍 Key Features

✅ Real-time Chatbot using BERT-based Intent Recognition.
✅ Deep Learning Model trained on an Intent JSON dataset.
✅ Optimized Text Processing & Tokenization.
✅ Accurate Intent Classification for diverse queries.
✅ Deployable on Web using Streamlit.

🚀 Technologies Used

Category	Tools & Libraries
Development	Python, Jupyter Notebook, Anaconda, VS Code
NLP Frameworks	Hugging Face Transformers, BERT
Machine Learning	TensorFlow, PyTorch
Data Processing	Pandas, NumPy
Deployment	Streamlit, Streamlit Cloud, HuggingFace

📷 Screenshots

Streamlit App - Chatbot Interface

Streamlit App - Chatbot Interface

Streamlit App - Chatbot Interface

Streamlit App - Chatbot Interface

🎯 Future Improvements

🔹 Expand dataset with more real-world conversations.
🔹 Integrate voice-based interaction using Speech Recognition.
🔹 Enhance context retention for long conversations.
🔹 Optimize model efficiency for faster response times.
🔹 Expanding chatbot capabilities with multilingual support.

📥 Installation & Setup

🔹 Clone the Repository

git clone https://github.com/Samarth4023/Shell-Internship-2.git
cd Shell-Internship-2

🔹 Install Required Dependencies

pip install -r requirements.txt

🔹 Run the Streamlit App

streamlit run app.py

📜 License

This project is open-source and free to use. Feel free to contribute!

📧 Contact

📌 Author: Samarth Pujari
📌 GitHub: Samarth4023
📌 LinkedIn: Samarth Pujari