Refine
Document Type
- Conference Proceeding (6)
- Article (3)
- Researchdata (3)
- Preprint (1)
- Workingpaper / Report (1)
Is part of the Bibliography
- yes (14)
Keywords
- Psychoakustik (4)
- Soundscape (3)
- Akustik (2)
- Datenanalyse (2)
- Glaubwürdigkeit (2)
- Maschinelles Lernen (2)
- Regressionsanalyse (2)
- Sprechen (2)
- Acoustic ecology (1)
- Acoustical properties (1)
Department/institution
Music listening is subjectively efficient in surrogating socioemotional needs during the Covid-19 pandemic, although underlying mechanisms remain unclear. We here challenge the generalizability of music-based coping by testing whether the insufficient satisfaction of distinct needs during the first lockdown entails enhanced listening to beat-based music with medium complexity (BBMMC) stimulating the cortico-striatal reward circuit and triggering oxytocinergic propagation. A Quasi-Bayesian mediation analysis on open-source data revealed that, across 11 European countries, more severe social distancing measures evoke a decline in experienced social provision (SP), in turn provoking BBMMC preferences. The latter was best predicted by the SP dimensions of social reward and attachment, suggesting that the partial overlap in neurophysiological processing at dopaminergic and oxytocinergic sites enables BBMMC to specifically alleviate these socioemotional needs, which has important implications for the development of mental health interventions. Finally, a stress-alleviating effect of BBMMC was found pointing towards HPA-activity modulation through music listening.
The set includes data related to the paper "The vibe of musical and social reward: Listening to beat-based music acts as a surrogate for socioemotional support during the Covid-19 pandemic"
*************** Information ***************
Dataset of the related text publication: https://doi.org/10.21203/rs.3.rs-2529116/v1
Soundscape studies vary considerably in study design, statistical methods, and model fit metrics used. Due to this confounding of data and methods, it is difficult to assess the suitability of statistical modelling techniques used in the literature. Therefore, five different methods and two performance metrics were applied to three existing soundscape datasets to model soundscape Pleasantness and Eventfulness based on seven acoustic and three sociodemographic predictors. Datasets differed in soundscape type (urban outdoor vs indoor), experimental setting (field- vs lab-based), size, and study design (site- vs person-centered). The fixed-effects and mixed-effects methods ranged from linear to nonlinear regression based on advanced machine learning approaches. Results showed that models performed better for Eventfulness than for Pleasantness in most cases, while performance as measured by the out-of-sample R2 was dependent on the total variance of the target, especially in both field studies with imbalanced targets and groups. Nonlinear methods consistently outperformed linear regression, with random forest and extreme gradient boosting performing particularly well, while the performance levels of all nonlinear methods remained comparable. Mixed-effects models provided a more generalized, albeit slightly smaller prediction performance when tested on unknown groups. Finally, this study motivates the use of cross-validation with special splitting for analyzing small imbalanced datasets.
The evaluation of listening experiments and studies on acoustics not only requires basic knowledge of statistics, but also poses challenges in the implementation of different methodologies. Especially for small sample sizes and peculiarities regarding hierarchical data structures, the need to individualize evaluation scripts arises. Therefore, we introduce flexcv, a powerful machine learning package for Python for evaluating various models on experimental tabular data, especially with small sample sizes. It supports random effects evaluation (including random slopes) for both linear and non-linear regressors, providing broad applicability to different experiments and research questions.flexcv quickly allows to perform nested cross-validation on a variety of models for comparison with each other. On the one hand, the implementation of a flexible interface simplifies the exchange of methods in the script, allowing researchers to change cross-validation methods without having to touch the actual cross-validation code. On the other hand, extensive online logging allows and simplifies the evaluation and experiment tracking along the process and different machines.
Exploring Album Dramaturgy: A Computational Music Analysis Across Pop Music History [Abstract]
(2024)
Background
A musical album can be considered a central artwork of an artist, and the order of its songs can influence the way the album is perceived by the listener. That is, purposeful sequencing can create an emotional journey and suspense (Lehne & Kölsch, 2015), tell a story or reinforce a particular concept or theme in the artist's music.
Aim(s)
This study thus aimed on examining the evolution of albums over the course of pop music history and on unveiling patterns (i.e. clusters) and temporal shifts in emotion- related album dramaturgy dependent on the cultural context (in terms of genres) by utilizing advanced computational methods.
Methods
We first created a dataset by scraping titles and artists of the weekly Top 10 albums (n = 14.394) from the Billboard.com charts from 1957 to 2022. We further extracted the emotion-related audio descriptors ‘valence’ and ‘energy’ from the Spotify Developer API for each track of the corresponding album, resulting in two-dimensional time series. We then employed Deep Time Series Embedding Clustering (DeTSEC; lenco et al., 2020) which can handle multivariate time series of different length. Finally, k-means clustering was applied on the embeddings produced by the DeTSEC, resulting in a 12 cluster solution.
Results
To facilitate the interpretation of our solution, we determined the most representative album of a cluster using the smallest Euclidean distance to its centroid. Preliminary time series analysis of the resulting 12 albums confirm different emotional trajectories of the respective clusters, for instance showing decreasing valence and arousal values over time for “I am up” by Young Thug (Cluster 6) or highest valence and arousal values at the beginning and the end of the album for“The New Classic” by Iggy Azalea (Cluster 12). Furthermore, Chi² tests revealed significant associations between an album’s cluster membership and both its year of release and its genre.
Discussion and Conclusion
The study constitutes a first step in shedding light on the evolution of albums in modern popular music since its emergence. It further provides an insight into how temporal dramaturgies in terms of the emotion-related arrangement of songs of albums differ across genres and undergo cultural trends and temporal shifts.
References
Ienco D, & Interdonato R (2020). Deep Multivariate Time Series Embedding Clustering via Attentive-Gated Autoencoder. Advances in Knowledge Discovery and Data Mining. 318–329
Lehne, M., & Koelsch, S. (2015). Toward a general psychological model of tension and suspense. Frontiers in Psychology, 6, 79.