A Python tool to:
- Find Definitions: Get precise meanings for complex words or phrases.
- Topic Information: Verify and cross-check information on specific topics. Done through cosine similarity and machine learning!
It helps any individual find verified information and definition of words or phrases in a simple, custom-built engine!
- I wanted to delve into an actual real-world application that I can devote my time on
- It was a way to accerlate my programming skills all levels (i.e., frontend & backend)
- Makes it easier for people to find verified information
- Project Lifespan: 4 months (Late December 2025 - Early April 2026; On and off between school and personal life)
Note on the Evolution of this Project: Initially, it was meant to be a traditional conversational AI Chatbot. However, during development, I decided to pivot towards a verification engine to focus on accuracy and implementation of semantic anlysis
- Month 1: Researching and attempting to build to a conversational chatbot
- Month 2: The pivot to focus on a high-precision verification engine
- Month 3: Training the Logistic Regression Model and implementing Cosine Similarity
- Month 4: UI development with Streamlit, debugging, and final deployment
- Python 3.13.2 - The programming language used for this project
- VS Code - My code editor
- Jupyter Notebook - Used for the Logistic Regression and Cosine Similairty Models
- Streamlit - For Web Interface
- Paste the link below on your browser:
- Semantic Analysis: Implementation of Machine Learning Models
- I learned to use Cosine Similarity and Logistic Regression
- API Integration: Connecting external data sources to code
- Object-Oriented Programming: Understanding the Class System in Python
- Full-Stack Integration: Combining frontend and backend
Note: This project is considered complete, but feel free to fork the repository. This is what I would do if I were to continue development
- Advanced NLP: Integrate enhanced NLP to imrpove the accuracy of semantic analysis beyond Cosine Similarity
- Expand Data Sources: Integrating more diverse APIs for enhanced verification capbilities
- User Accounts: Adding a history feature so users can save and revisit their previous verified information
- RAG Implementation: Allows the bot to generate natural language answers grounded strictly in the verified data it retrives
- Mobile Otimization: Further refining the Streamlit UI for a better mobile experience
- Various Tutorials on Youtube that helped with basic ML and Cosine Similairty
- Google Gemini: Used as a debugging assistant to help interpret terminal error messages and explain the math behind Cosine Similarity and Logistic Regression
- mahdia1-ops: For giving me the idea, helping out with the data saving, and helping me clean up the method on finding definitions
