C4DM PhD researchers Vjosa Preniqi and Iacopo Ghinassi, funded via the DAME CDT, have developed a new web app that automatically predicts moral values in text and music lyrics. Users can upload CSV or Excel files containing text or lyrics, and the app will assign probabilities for 10 moral categories. It is particularly effective for social media posts and music lyrics. This new tool provides a simple, user-friendly interface for analysing how language reflects moral beliefs, offering useful insights for research in music and media analysis. Try it out!
The app is based on models from two recent papers:
- MoralBERT: A Fine-Tuned Language Model for Capturing Moral Values in Social Discussions (ACM GoodIT 2024)
- Automatic Detection of Moral Values in Music Lyrics (ISMIR 2024)
This work is a collaboration between C4DM and ISI Foundation.