Skip to content

ananyag309/Bank_of_Baroda_Hackathon_Project

 
 

Repository files navigation

Bank of Baroda Chatbot Project

Project Overview

This project is developed by the team Ctrl Freaks for the Bank of Baroda Hackathon. Our goal is to create an intelligent chatbot that can assist customers with common queries and provide useful financial information.

Features

  • Greeting and Farewell Responses: The chatbot can handle greetings and farewells, providing a friendly interaction experience.
  • Query Responses: The chatbot can answer questions about the existence of God, current inflation rates, income levels, and more.
  • Financial Assistance: The chatbot offers responses to various financial-related questions, including savings management, financial goals, risk attitudes, and more.

Technologies Used

  • Python: The core programming language used for the chatbot development.
  • TensorFlow: For building and training the neural network model.
  • Natural Language Toolkit (nltk): For text processing and tokenization.
  • Keras: For building and saving the model.
  • JSON: For handling the intents and responses.
  • Pickle: For saving and loading processed data.

Project Structure

  • intents.json: Contains the intents, patterns, responses, and context of the chatbot.
  • chatbot_model.h5: The trained model file.
  • words.pkl: Pickle file for storing the processed words.
  • classes.pkl: Pickle file for storing the processed classes.

Installation

  1. Clone the repository:

    git clone https://github.com/your-repo/bob-chatbot.git
    cd bob-chatbot
  2. Install the required dependencies:

    pip install numpy nltk tensorflow keras
  3. Download NLTK data:

    import nltk
    nltk.download('punkt')
    nltk.download('wordnet')
  4. Run the chatbot:

    python chatbot.py

What We Have Done So Far

  • Data Preparation: Collected and processed the data, tokenized the patterns, and created training datasets.
  • Model Training: Developed and trained a neural network model using TensorFlow and Keras.
  • Response Handling: Implemented functions to predict the class of the input sentence and retrieve appropriate responses.

Current Status:

Next Steps

  • Improve Model Accuracy: Enhance the model by fine-tuning hyperparameters and expanding the dataset.
  • Add More Intents: Include more intents to cover a broader range of customer queries.
  • Integrate with Bank Systems: Connect the chatbot to Bank of Baroda's systems for real-time data retrieval and transactions.
  • Deploy the Chatbot: Host the chatbot on a server and create a user-friendly interface for customers.

Team - Ctrl Freaks


Ananya Gupta

Neha Saini

Peehu Mishra

Aditi Jain

License

This project is licensed under the MIT License. See the LICENSE file for details.

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%