Simple Streamlit dashboard created using Official Youtube API and Spacy NLP similarity to search for similar youtube channels you might enjoy, using a youtube link as input.
- Set-up Youtube project to obtain personal API Key. See official documentation for details
- Note: Take note of daily API call quota limits (see official limits for details).
-
Activate virtual environment in the directory of choice and install the necessary libraries outlined in requirements.txt and the required NLP pipeline for spacy dependecy.
pip install -r requirements.txt python -m spacy download en_core_web_lg
-
Activate your virtual environment of choice and run the streamlit script dashboard.py
streamlit run dashboard.py
-
The Dashboard would be running on local host (Port: 8501) by default. Open the web browser and enter the corresponding localhost address (http://127.0.0.1:8501/) to view the Dashboard.
-
To start using the dashboard, fill the following fields before selecting the 'Submit' button:
- API Key: Obtain in Pre-requisite step 1
- Search Input: Youtube input link (Example: https://www.youtube.com/watch?v=a1b2c3d4e5f)
- Max Search Results: Maximum number of results to be returned
- Similarity Filter Threshold: Channels with similarity score of input threshold or higher will be returned after search. (Higher threshold corresponds to more similar channels returned)
