Analyzing the Impact of Audio Features on Popularity in Global Music Consumption using Machine Learning: A Spotify Case Study

Abstract

The rapid growth of music streaming platforms such as Spotify has transformed how the audience consume music globally. This research analyses the influence of audio features on the song popularity in global music consumption using machine learning utilizing a Spotify dataset of around 2000 records. The audio attributes used for analyzation are acousticness, danceability, energy, instrumentalness, liveness, loudness, speechiness, valence, and tempo of the audio track along with the track popularity metrics, genres and release dates. The research is aimed to identify patterns and factors that contribute to a song's success. The research significance is that its potential to bridge the gap between the data-driven insights and global music consumption. Logistic Regression, Random Forest and Artificial Neural Networks are employed to predict song popularity and evaluate the importance of individual features driving audience engagement. The methodology involves data gathering, preprocessing, feature engineering, algorithm selection, model training and model evaluation. Data preprocessing phase involves handling missing values, normalizing numerical features and encoding categorical variables to ensure consistency. Feature engineering includes selecting key audio attributes, creating new derived features and applying dimensionality reduction techniques to enhance model performance. For each algorithm, precision, recall, F1-score and accuracy values were computed. Accuracy values for Logistic Regression, Random Forest and ANN are 93%, 92% and 90% respectively. The findings offer insights into how specific audio characteristics influence listener preferences and provide a model which supports the artists, producers, and streaming platforms to optimize song releases for high popularity ratings. The model incorporates global trends by analyzing playlist genres and listener engagement to reveal how audio features influence music consumption. This research contributes to the growing music industry and highlights the potential of machine learning to reveal patterns in global music consumption. The results help stakeholders make informed decisions in music production, curation, and marketing.