A New Audio Approach Based on User Preferences Analysis to Enhance Music Recommendations
Bibliographic record
Abstract
With the recent upsurge in music consumption, music recommendation systems have gained substantial prominence.Platforms like Spotify are increasingly relied upon by users for curated music, underscoring the need for improved recommendation algorithms.While the analysis of user preferences and historical listening behaviours has conventionally been employed to tailor recommendations, these techniques are often restricted to examining textual data, such as lyrics and titles, thereby potentially limiting the effectiveness of the recommendations.The current study proposes a novel approach that extends beyond textual analysis to investigate the audio aspect of music, which directly influences listeners' emotions.This exploration encompasses the feature extraction and selection phases based on multi-models, contributing to robustness and interpretability, especially when contending with noise generated by the audio signal.Three distinct strategies for feature extraction and selection were incorporated, focusing on musical characteristics such as speed, rhythm, tonality, and signal changes.These strategies employed Librosa, PyAudio analyses, and Convolutional Neural Networks (CNNs) using the VGG16 model.Subsequently, features were classified to assess their efficacy and provide a preliminary evaluation of the proposed recommendation system.The system's personalisation was achieved by enabling users to select a piece of music, from which their preferences were extracted.The efficacy of this approach was validated through extensive experiments using the GTZAN dataset, comprising 10 distinct music genres with 100 audio files lasting 30 seconds each.Findings suggest that CNNs present a reliable method for generating personalised music recommendations, particularly for users with preferences for similar artists or diverse genres.Conversely, for users favouring a specific genre, Librosa appeared to provide a more effective means of achieving optimal recommendation accuracy.Therefore, this study illuminates new pathways for music analysis and classification, with the ultimate goal of enhancing understanding of the auditory world and improving the music recommendation experience for users.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.007 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".