Text-to-Emoji

A powerful, local semantic search engine that finds the perfect emoji using natural language. Built with the Streamlit framework and Sentence Transformers, it allows users to search for emojis conceptually (e.g., "victory" → ✌️, 🏆, 🎆) rather than just by keyword. It uses a persistent local vector database for fast, efficient queries.

🚀 Features

• Natural Language Search: Type complex queries like "feeling loved" or "space travel" to find matching emojis.
• Semantic Understanding: Uses a transformer model to understand the meaning behind emojis, not just their names.
• Blazing Fast: Powered by sqlite-vec for millisecond-level vector similarity search.
• Persistent Index: efficient local storage using SQLite; indexes thousands of emojis once and reuses the database.
• Modern UI: Clean, responsive interface built with Streamlit, featuring gradient cards and match percentage scores.

� Screenshots

🛠️ Tech Stack

Core

• Python: The primary programming language.
• Sentence Transformers: all-MiniLM-L6-v2 model for generating dense vector embeddings.

Frontend

• Streamlit: For building the interactive web application.

Database

• SQLite: Reliable local database engine.
• sqlite-vec: Vector search extension for SQLite.

Data

• Requests: For fetching the latest emoji definitions from Unicode.org.

📋 Prerequisites

Before installation, ensure you have the following:

• Python 3.12+
• pip

🔧 Installation

1. Clone the repository

   git clone <repository-url>
   cd text-to-emoji

2. Install dependencies

   pip install -r requirements.txt

   (Note: Ensure sqlite-vec is installed correctly for your platform)

🚀 Running the App

Start the Application

streamlit run main.py

Accessible at http://localhost:8501

On the first run, the app will download the embedding model and index the emoji dataset. This is a one-time process.