A Python tool for semantic search in YouTube video transcripts.
- Search for specific phrases in YouTube video transcripts
- Semantic matching to find similar phrases regardless of exact wording
- Configurable similarity threshold
- Support for both pip and pipenv workflows
pip install -r requirements.txt
pip install pipenv
pipenv install
pipenv shell
python transcript.py <youtube_url> "<search_phrase>" [options]
Parameter | Description | Required | Default |
---|---|---|---|
youtube_url | Full URL of the YouTube video | Yes | - |
search_phrase | Text to search for in the transcript | Yes | - |
--threshold | Minimum similarity threshold (0-100) | No | 80 |
Search for an exact phrase:
python transcript.py "https://www.youtube.com/watch?v=hX4KgFNuwZ8" "get to 10 million" --threshold 85
Search with semantic matching (finds similar meanings):
python transcript.py "https://www.youtube.com/watch?v=hX4KgFNuwZ8" "get to ten milion" --threshold 85
- Python 3.7 or higher
- See
requirements.txt
for package dependencies