Implementasi Deep Learning Menggunakan Kerangka Kerja Cross-Industry Standard Process For Data Mining (CRISP- DM) Pada Peminat Genre Musik Di Aplikasi Spotify
Kata Kunci:
Deep Learning, CRISP-DM, Spotify, K-Nearest Neighbor, Naive Bayes, Peminat GenreAbstrak
Penelitian ini bertujuan memetakan peminat genre musik pada tingkat kumpulan lagu Spotify melalui implementasi deep learning menggunakan kerangka kerja Cross-Industry Standard Process for Data Mining (CRISP-DM). Dataset berasal dari rekomendasi halaman utama Spotify yang dikompilasi ke dalam satu playlist penelitian pada tahun 2024, kemudian diekspor melalui Exportify. Setelah pembersihan data, diperoleh 7.858 track dengan 15 fitur audio dan metadata numerik. Label genre dibentuk secara bertingkat dari genre_awal, distandardisasi menjadi genre_utama, lalu dikelompokkan ke delapan meta_genre, yaitu Rock/Metal/Punk, Pop/Indo/Malay, Other, Urban, Country/Folk, EDM/Dance, Latin, dan Jazz/Classical. Pemodelan dilakukan dengan k-Nearest Neighbor (k-NN), Gaussian Naive Bayes, dan model gabungan berbasis rata-rata probabilitas. Evaluasi menggunakan accuracy, precision, recall, dan weighted F1-score pada data validasi dan data uji, disertai pengujian kestabilan melalui beberapa pengulangan stratified split. Hasil menunjukkan bahwa k-NN memberikan kinerja terbaik pada data uji dengan accuracy 65,39% dan weighted F1-score 62,61%. Model gabungan mencapai weighted precision tertinggi pada validasi sebesar 67,45%, tetapi masih sensitif terhadap variasi pembagian data. Deployment pada seluruh dataset memperlihatkan dominasi Rock/Metal/Punk sebesar 62,4% dan Pop/Indo/Malay sebesar 11,7%. Temuan ini menunjukkan bahwa CRISP-DM mampu menata proses klasifikasi genre secara sistematis, sementara hasil deployment dapat digunakan untuk membaca peminat genre pada tingkat kumpulan lagu, bukan pada tingkat pengguna individual.
Referensi
[1] M. Schedl, E. Gomez, and J. Urbano, “Music Information Retrieval: Recent Developments and Applications,” Foundations and Trends in Information Retrieval, vol. 8, no. 2-3, pp. 127-261, 2014.
[2] D. Duman, P. Neto, A. Mavrolampados, P. Toiviainen, and G. Luck, “Music we move to: Spotify audio features and reasons for listening,” PLOS ONE, vol. 17, no. 9, e0275228, 2022.
[3] F. Khan et al., “Effect of Feature Selection on the Accuracy of Music Popularity Classification Using Machine Learning Algorithms,” Electronics, vol. 11, no. 21, 3518, 2022.
[4] J. Ramirez and M. J. Flores, “Machine learning for music genre: Multifaceted review and experimentation with audioset,” Journal of Intelligent Information Systems, vol. 55, no. 3, pp. 469-499, 2020.
[5] G. Ru, X. Zhang, J. Wang, N. Cheng, and J. Xiao, “Improving Music Genre Classification from Multi-Modal Properties of Music and Genre Correlations Perspective,” arXiv, 2023.
[6] C. Schroer, F. Kruse, and J. M. Gomez, “A Systematic Literature Review on Applying CRISP-DM Process Model,” Procedia Computer Science, vol. 181, pp. 526-534, 2021.
[7] R. Wirth and J. Hipp, “CRISP-DM: Towards a Standard Process Model for Data Mining.”
[8] S. Berkovsky, I. Cantador, and D. Tikk, Collaborative Recommendations: Algorithms, Practical Challenges and Applications. World Scientific, 2018.
[9] C. Bauer and M. Schedl, “Global and country-specific mainstreaminess measures: Definitions, analysis, and usage for improving personalized music recommendation systems,” PLOS ONE, vol. 14, no. 6, e0217389, 2019.
[10] D. Sanchez-Moreno, Y. Zheng, and M. N. Moreno-Garcia, “Time-Aware Music Recommender Systems: Modeling the Evolution of Implicit User Preferences and User Listening Habits in a Collaborative Filtering Approach,” Applied Sciences, vol. 10, no. 15, 5324, 2020.
[11] A. Caronongan and R. Cabredo, “Modeling User Music Preference Through Usage Scoring and User Listening Behavior for Generating Preferred Playlists,” in Lecture Notes in Computer Science, vol. 9935, p. 73, 2016.
[12] T. Pratiwi, A. Sunyoto, and D. Ariatmanto, “Music Genre Classification Using K-Nearest Neighbor and Mel-Frequency Cepstral Coefficients,” Sinkron, vol. 8, no. 2, pp. 861-867, 2024.
[13] A. Syah, B. Yuliadi, and R. Sahara, “Music Genre Classification using Naive Bayes Algorithm,” International Journal of Computer Trends and Technology, vol. 62, no. 1, pp. 50-57, 2018.
[14] I. M. De Diego, A. R. Redondo, R. R. Fernandez, J. Navarro, and J. M. Moguerza, “General Performance Score for classification problems,” Applied Intelligence, vol. 52, no. 10, pp. 12049-12063, 2022.
[15] D. J. Hand, P. Christen, and N. Kirielle, “F*: An interpretable transformation of the F-measure,” Machine Learning, vol. 110, no. 3, pp. 451-456, 2021.
[16] B. Vrigazova, “The Proportion for Splitting Data into Training and Test Set for the Bootstrap in Classification Problems,” Business Systems Research Journal, vol. 12, no. 1, pp. 228-242, 2021.
