Data Visualization and
Query using LLM Models
E xplore how a robust football database, natural language processing,
and interactive data visualization can empower users to uncover
valuable insights from sports data.
TEAM - A
Aryan Jadhav
Akshay Reddy
Harsha Sharma
Kartikyae Dhapola
Keshavi Sharma
Aditi Satsangi
Tamanna
Mentor Name : Bhargava Sai Reddy
Date : 26/7/24
Introduction
1 Objective 2 Modules Overview
Database Creation, SQL Query Generation using
To create a system that enables users to query a
LLM, Data Visualization with Streamlit,
football database using natural language and visualize
Documentation.
the results.
SQL Query Generation using LLM
Approach Process Outcome
Integrate OpenAI API with Python Develop a prompt to convert user A system where users can input
to enable natural language to SQL input into SQL queries, test and questions in plain English and
query conversion. refine for accuracy and relevance. receive SQL queries as output.
Data Visualization with Stream lit
Tool Used
Streamlit, a powerful open-source app framework for building data-driven
web applications.
Visualization Types
Line charts, bar charts, histograms, and other interactive visualizations.
Process
Connect SQL query results to Streamlit and implement dynamic, user-
friendly charts.
Project Overview
Objective Key Features
The main objective of this project is to simplify data Our system offers natural language query interpretation,
extraction and visualization for users who may not have secure database connection, and dynamic data
technical expertise in SQL or other query languages. visualization through Streamlit.
Project Deliverables
1 Developed System
We have developed a fully functional system that interprets
natural language queries, translates them into SQL, and
visualizes the results dynamically.
2 Documentation
Comprehensive documentation has been prepared,
detailing the system architecture, code explanations, and
user instructions.
3 Future Enhancements
Future enhancements may include support for more complex
queries, integration with additional databases, and enhanced
visualization options.
Salient features of our project
• Natural Language Query Processing
• Advanced NLP Integration
• Interactive Data Visualization
• Secure Database Connectivity
• Customizable Visualizations
Demonstration
User Input
User enters a natural language question about football
statistics.
SQL Generation
The system generates the corresponding SQL query
using the LLM model.
Visualization
The query results are displayed in an interactive
Streamlit dashboard.
Challenges and Solutions
Challenges Solutions
E nsuring accurate SQL query generation, handling Iterative testing and prompt refinement, leveraging
diverse and complex queries, efficient data visualization Streamlit's features for better user experience,
and user interaction. continuous feedback and improvements.
Conclusion
1 Summary and Key Outcomes
• Functional System: Successfully built a system that integrates
database management, natural language processing, and data
visualization.
• User Engagement: Created a user-friendly interface that allows
non-technical users to query and visualize data effectively.
2 Success in Meeting Objectives
• Effective Database Management: The project successfully
established a well-structured and Accurate database, meeting
the objective of efficient data management.
• Accurate SQL Query Generation: The implementation of GPT-
3.5 for natural language processing resulted in precise and
reliable SQL query generation.
Thank You
We hope this presentation has provided a glimpse into the exciting world of data-driven insights and natural
language-powered applications.
