Introducing EEG Analyses to Help Personal Music Preference Prediction

TL;DR

This study investigates using portable dry-electrode EEG signals as explicit feedback to improve personal music preference prediction. By collecting EEG, mood, and preference labels during music listening, the authors analyze relationships between neural activity, mood states, and user preferences, and show that PSD features from EEG improve rating prediction and preference classification beyond baseline features. The work demonstrates the feasibility of integrating EEG into a music recommender in everyday contexts and discusses practical considerations, privacy, and limitations of dry-electrode devices. The findings suggest a broader potential for EEG-based explicit feedback in personalized recommendation tasks beyond music.

Abstract

Nowadays, personalized recommender systems play an increasingly important role in music scenarios in our daily life with the preference prediction ability. However, existing methods mainly rely on users' implicit feedback (e.g., click, dwell time) which ignores the detailed user experience. This paper introduces Electroencephalography (EEG) signals to personal music preferences as a basis for the personalized recommender system. To realize collection in daily life, we use a dry-electrodes portable device to collect data. We perform a user study where participants listen to music and record preferences and moods. Meanwhile, EEG signals are collected with a portable device. Analysis of the collected data indicates a significant relationship between music preference, mood, and EEG signals. Furthermore, we conduct experiments to predict personalized music preference with the features of EEG signals. Experiments show significant improvement in rating prediction and preference classification with the help of EEG. Our work demonstrates the possibility of introducing EEG signals in personal music preference with portable devices. Moreover, our approach is not restricted to the music scenario, and the EEG signals as explicit feedback can be used in personalized recommendation tasks.

Figures (11)

• Figure 1: User study procedure and the user-computer interface of the experiments.
• Figure 2: Two-dimensional description map of Thayer mood, where the X-axis represents valence, and the Y-axis represents arousal.
• Figure 3: The portable EEG headset deployed in our lab study with the mark of electrodes and their corresponding sensor locations. Red circles are the recorded electrodes, and the blue circle is the reference electrode.
• Figure 4: Time series data before and after removing low-frequency drift for a sample.
• Figure 5: The distribution of mood (valence and arousal) in each preference (in terms of rating) and the relationship between mood and preference. The colors represent preferences(in terms of ratings). P stands for $p$ value from the ANOVA test between 5 groups(rating 1-5).