Textube

This project is a web application that allows users to perform searches within media files (audio or video) transcriptions. The search functionality supports fuzzy matching, stemming, and exact phrase matching.

Features

• Extract transcriptions from YouTube videos or uploaded media files.
• Store transcriptions and their timestamps in Elasticsearch.
• Perform advanced text search with:
  • Fuzzy matching
  • Stemming and stop word removal
  • Exact phrase matching
• Highlight matching words in search results.

Technologies Used

• Backend: Flask
• Search Engine: Elasticsearch
• Frontend: HTML, CSS, JavaScript
• Transcription: Whisper

Installation and Setup

Prerequisites

Ensure you have the following installed:

• Python 3.10
• Elasticsearch
• Flask

Steps to Run the Project

1. Clone the repository:

   git clone https://github.com/your-username/transcription-search-app.git
   cd transcription-search-app

2. Install dependencies:

   pip install -r requirements.txt

3. Start Elasticsearch (if not running already).
4. Run the Flask application:

   python app.py

5. Open your browser and navigate to http://localhost:5000.

API Endpoints

POST /process

• Description: Processes a video link or uploaded file and indexes its transcription in Elasticsearch.
• Parameters:
  • videoLink (string): YouTube video URL (optional)
  • mp4Upload (file): Uploaded media file (optional)
  • sentence (string): Sentence to search within transcriptions

GET /search

• Description: Searches for a sentence in the indexed transcriptions.
• Parameters:
  • query (string): The search term

Elasticsearch Indexing

The project uses a custom Elasticsearch index with analyzers:

• simple_analyzer: Lowercase only.
• custom_analyzer: Lowercase, stopword removal, and stemming.

Future Improvements

• Migrate the frontend to React for a better user experience.
• Implement a real-time transcription feature.
• Enhance search results with word-level timestamps.

License

This project is licensed under the MIT License.

Author: Youxise