Predicting the Preference for Sad Music: The Role of Gender, Personality, and Audio Features

Abstract

The "tragedy paradox" of music, avoiding experiencing negative emotions but enjoying the sadness portrayed in music, has attracted a great deal of academic attention in recent decades. Combining experimental psychology research methods and machine learning techniques, this study (a) investigated the effects of gender and Big Five personality factors on the preference for sad music in the Chinese social environment and (b) constructed sad music preference prediction models using audio features and individual features as inputs. Statistical analysis found that males have a greater preference for sad music than females do, and that gender and the extraversion factor are involved in significant two-way interactions. The best-performing random forest regression shows a low predictive effect on the preference for sad music (R 2 = 0.138), providing references for music recommendation systems. Finally, the importance-based model interpretation feature reveals that, in addition to the same music inputs (audio features), the perceived relaxation and happiness of music play an important role in the prediction of sad music preferences.

RETRACTED: The approaches and methods of music psychology in the relationship between music emotion and cognition in music teaching activities

“…Junior high school students are exposed to a lot of music and have formed certain esthetic standards. Unlike primary school students, they are not easy to accept the esthetic content arranged by teachers ( Xu et al, 2021 ). Moreno, M. et al pointed out that in music appreciation activities, esthetic subjects always judge esthetic objects with certain esthetic standards.…”

Integrating Individual Factors to Construct Recognition Models of Consumer Fraud Victimization

“…For the machine learning algorithm, this work applied the random forest classification (RFC) algorithm. As RFC has shown good performance in classification tasks [ 48 , 49 ], we can easily interpret the constructed RFC models by calculating the feature importance [ 50 , 51 , 52 , 53 , 54 ]. The predictive effect of each model was evaluated by the tenfold cross-validation technique.…”

“…The predictive effect of each model was evaluated by the tenfold cross-validation technique. The tenfold cross-validation technique uses 90% of the data as training data to train the models and the remaining instances as testing data, and this procedure is repeated ten times [ 50 ]. Finally, the prediction accuracy of each classifier was measured using precision, recall, and F1 values as follows [ 55 ]: Precision = TP/(TP + FP) Recall = TP/(TP + FN) F1 = 2 × Precision × Recall/(Precision + Recall) where TP (true positive) is the number of positive samples predicted by the classifier as positive; FP (false positive) is the number of negative samples predicted by the classifier as positive; and FN (false negative) is the number of positive samples predicted by the classifier as negative.…”

Graph-Based Attentive Sequential Model With Metadata for Music Recommendation

“…Lin et al [16] focused on heterogeneous information obtained on music media streaming platforms and designed knowledgebased neural networks utilizing graphic, textual, and visual data. When it comes to the preference for sad music, Xu et al [17] investigated the relevance between personalities and audio features from both psychological and informatics perspectives.…”