/DataMoshpit2025

Demo for DataMoshpit - Psycho-social Metal šŸ¤˜ Analysis

Primary LanguagePython

DataMoshpit2025

Demo for DataMoshpit - Psycho-social Metal šŸ¤˜ Analysis

Theory behind

Trait data: There is no scientific dataset that has measured Big Five or psychosocial values for each band individually. What we can do is approximate band profiles from their subgenre tags, because the psychological research gives us associations at the genre family level (e.g., ā€œextreme metalā€ ā†’ higher nonconformity/uniqueness, ā€œprogressiveā€ ā†’ higher openness, etc.).

ToDO:

  1. Loads the full band roster (e.g. from Kaggleā€™s Metal Archives dataset URL: https://www.kaggle.com/datasets/guimacrlh/every-metal-archives-band-october-2024/data or via API : https://www.metal-api.dev/index.html )
  2. Normalizes the genre text.
  3. Uses a subgenre ā†’ trait mapping dictionary (like the one weā€™ve started, but expanded).
  4. Outputs a CSV: Band, Country, Genre, Openness, Nonconformity, Authority_skepticism, Extraversion, Emotion_regulation, Sensation_seeking
Band Genre Openness Nonconformity Authority skepticism Extraversion Emotion regulation Sensation-seeking
Slayer Thrash Metal 0.68 0.60 0.60 0.42 0.60 0.70
Opeth Progressive Metal 0.85 0.60 0.50 0.40 0.60 0.60
Nightwish Symphonic Metal 0.70 0.50 0.30 0.60 0.55 0.50
Behemoth Blackened Death 0.70 0.75 0.70 0.30 0.65 0.75
Sabaton Power Metal 0.65 0.45 0.30 0.70 0.50 0.55

Local run

  1. Questionnaire on app run metal_recommender_app.py and make sure to have installed streamlit:
pip install streamlit 
pip install pandas 
pip install numpy
streamlit run metal_recommender_app.py
  1. go to folder and run:
streamlit run metal_recommender_app.py

open URL: http://localhost:8501/

Deployment on Azure Web app (stat)

  1. Have Resource group created and Web App created
  2. Run Github Actions and change YAML file
  3. Have permissions on on resource group added
  4. Add role assignment (role = Contributor) for your Web app