TROMPA-MER: an open dataset for personalized Music Emotion Recognition

Abstract

We present a platform and a dataset to help research on Music Emotion Recognition (MER). We developed the Music Enthusiasts platform aiming to improve the gathering and analysis of the so-called “ground truth” needed as input to MER systems. Firstly, our platform involves engaging participants using citizen science strategies and generate music emotion annotations -- the platform presents didactic information and musical recommendations as incentivization, and collects data regarding demographics, mood, and language from each participant. Participants annotated each music excerpt with single free-text emotion words (in native language), distinct forced-choice emotion categories, preference, and familiarity. Additionally, participants stated the reasons for each annotation -- including those distinctive of emotion perception and emotion induction. Secondly, our dataset was created for personalized MER and contains information from 181 participants, 4721 annotations, and 1161 music excerpts. To showcase the use of the dataset, we present a methodology for personalization of MER models based on active learning. The experiments show evidence that using the judgment of the crowd as prior knowledge for active learning allows for more effective personalization of MER systems for this particular dataset. Our dataset is publicly available and we invite researchers to use it for testing MER systems.

AI for Musical Discovery

“…MER often depends on datasets and prediction targets derived from simplistic taxonomies. 30 Implicit learning from textual associations may instill biases from datasets, due to limitations in how emotion is often discussed and the need for more information than language alone for representing rich emotional concepts. 31 Such techniques lack grounding in the human experiences, embodiment, and enculturation that gives rise to musical emotional fluency, and may implicitly encode biases in music-emotion connections.…”