Title

基於隱含情境因素之行動音樂推薦系統

Translated Titles

A Mobile Music Recommender System Based on Latent Context Factors

Authors

翁世昕

Key Words

音樂推薦 ; 情境感知推薦系統 ; 隱含情境因素 ; 音樂資訊檢索 ; Music Recommender System ; Context-Aware Recommender System ; Latent Context Factor ; Music Information Retrieval

PublicationName

成功大學資訊管理研究所學位論文

Volume or Term/Year and Month of Publication

2016年

Academic Degree Category

碩士

Advisor

劉任修

Content Language

繁體中文

Chinese Abstract

網路音樂平台快速發展使得使用者回饋資訊更容易的被蒐集,造成了資訊過載的現象,音樂推薦系統的重要性日益漸增。在協同過濾法(Collaborative Filtering)的推薦演算法中,基於模型式方法(Model-Based Approach)除了預測階段速度快,亦擁有不錯的推薦結果, 然而當新歌曲進入系統後,系統並沒有足夠的資料進行分析,造成了冷開始的問題(Cold Start)。 不同於協同過濾,內容基礎式過濾法(Content-Based Filtering) 則是加入額外對於歌曲音訊內容的描述,有能力處理由新歌曲所造成的冷開始問題,但推薦結果通常較前者差。 為了同時具備對於兩種情況的處理能力,混合式過濾法 (Hybrid Methods)結合了前述方法滿足了這樣的需求。 本論文延續同時考量使用者回饋資料和音訊內容的音樂推薦的研究,加入環境資料設計了行動情境感知式推薦系統(Context-Aware Recommender System),能夠依據使用者當前所在的環境給予推薦結果,基於假設使用者會在不同環境下有不同喜好,將提高推薦結果的準確性。 一旦本推薦模型能夠分析了解使用者在不同環境下的喜好,就可以自動建立播放清單,在使用者不能手動選擇播放的歌曲時仍能夠藉由此推薦模型聆聽當下喜好的歌曲。 實驗資料集是來自於手機中的程式自動蒐集使用者回饋資訊和環境資料等等,目的是避免影響使用者實際聆聽音樂的習慣,以確保資料的真實性。 本論文所提出之隱含情境因子推薦模型(Latent Context Factors-Recommender Model )將同時考量使用者回饋資訊、歌曲內容和對應的情境資料,除了有能力依照當前環境給予推薦外,亦保有基於模型式方法的特性。 而根據本論文的實驗結果顯示本論文提出的推薦模型相較於Yoshii et al. (2008)和Su et al. (2010)所提出的方法為佳,證明本模型擁有一定水準的推薦正確率,且在模型中考環境相較於未考慮環境有些許的改善。

English Abstract

In this thesis, a new music recommender system called the Latent Context Factors Recommender Model is proposed. The goal is to learn which kind of music users like in different contexts. To reach the goal, we implement a smart-phone application to collect various data regarding music listening habits. Users who agree to participate in this experiment do not need to manually report. Instead, our program automatically collects the required data. The collected record includes the user's ID, the title of the audio file, and the environmental attributes measured by the smart-phone's sensors. These sensory attributes are helpful for inferring which environment the user was in while listening to the music. Apart from the environment, audio content is also important to our study. We analyze audio files and timbre attributes to characterize the content of the audio files. Based on the collected data, our model is able to estimate occurrence probabilities of the users in different latent contexts, and to learn each user's preferences in those contexts. According to the results of our analysis, recommendations are provided to users. Because users may select songs according to what they are doing, estimating a user's probability of appearing in a given context is helpful to learning their preferences. Results of several measures show that our recommender model has greater accuracy than other models. However, the difference between the results with and without considering environmental attributes is not significant.

Topic Category 管理學院 > 資訊管理研究所
社會科學 > 管理學
Reference
  1. Ashley-Dejo, E., Ngwira, S., & Zuva, T. (2015). A survey of context-aware recommender system and services. In International conference on computing, communication and security (pp. 1–6).
    連結:
  2. Baltrunas, L., Ludwig, B., Peer, S., & Ricci, F. (2012). Context relevance assessment and exploitation in mobile recommender systems. Personal and Ubiquitous Computing, 16(5), 507–526.
    連結:
  3. Bhatia, N., et al. (2010). Survey of nearest neighbor techniques. International Journal of Computer Science and Information Security, 8(2), 302–305.
    連結:
  4. Celma, O. (2010). Music recommendation. Berlin: German:Springer.
    連結:
  5. Chen, Z.-S., Jang, J.-S. R., & Lee, C.-H. (2011). A kernel framework for content-based artist recommendation system in music. IEEE Transactions on Multimedia, 13(6), 1371–1380.
    連結:
  6. De Lathauwer, L., De Moor, B., & Vandewalle, J. (2000). A multilinear singular value decomposition. SIAM journal on Matrix Analysis and Applications, 21(4), 1253–1278.
    連結:
  7. Ding, C., Li, T., & Peng, W. (2006). Nonnegative matrix factorization and probabilistic latent semantic indexing: Equivalence chi-square statistic, and a hybrid method. In Proceedings of the national conference on artificial intelligence (Vol. 21, p. 342).
    連結:
  8. Downie, J. S. (2003). Music information retrieval. Annual review of information science and technology, 37(1), 295–340.
    連結:
  9. Iwahama, K., Hijikata, Y., & Nishida, S. (2004). Content-based filtering system for music data. In 2004 international symposium on applications and the internet workshops (pp. 480–487).
    連結:
  10. Koren, Y., Bell, R., & Volinsky, C. (2009). Matrix factorization techniques for recommender systems. IEEE Computer Society, 42(8), 30–37.
    連結:
  11. Kumar Bokde, D., Girase, S., & Mukhopadhyay, D. (2014). Role of matrix factorization model in collaborative filtering algorithm: A survey. International Journal of Advance Foundation and Research in Computer, 1, 111-118.
    連結:
  12. Logan, B., & Salomon, A. (2001). A music similarity function based on signal analysis. In The ieee international conference on multimedia (p. 190).
    連結:
  13. Reynolds, D. A., Quatieri, T. F., & Dunn, R. B. (2000). Speaker verification using adapted gaussian mixture models. Digital signal processing, 10(1), 19–41.
    連結:
  14. Ricci, F., Rokach, L., & Shapira, B. (2011). Recommender systems handbook. Berlin: German:Springer.
    連結:
  15. Shao, B., Wang, D., Li, T., & Ogihara, M. (2009). Music recommendation based on acoustic features and user access patterns. IEEE Transactions on Audio, Speech, and Language Processing, 17(8), 1602–1611.
    連結:
  16. Su, J.-H., Yeh, H.-H., Yu, P. S., & Tseng, V. S. (2010). Music recommendation using content and context information mining. IEEE Intelligent Systems, 25(1), 16–26.
    連結:
  17. Wang, X., & Wang, Y. (2014). Improving content-based and hybrid music recommendation using deep learning. In Proceedings of the acm international inproceedings on multimedia (pp. 627–636).
    連結:
  18. Yoshii, K., Goto, M., Komatani, K., Ogata, T., & Okuno, H. G. (2008). An efficient hybrid music recommender system using an incrementally trainable probabilistic generative model. IEEE Transactions on Audio, Speech, and Language Processing, 16(2), 435–447.
    連結:
  19. Yujie, Z., & Licai, W. (2010). Some challenges for context-aware recommender systems. In International conference on computer science and education (pp. 362–365).
    連結:
  20. Baltrunas, L., Kaminskas, M., Ludwig, B., Moling, O., Ricci, F., Aydin, A., . . .Schwaiger, R. (2011). Incarmusic: Context-aware music recommendations in a car. In Ec-web (Vol. 11, pp. 89–100).
  21. Bennett, J., & Lanning, S. (2007). The netflix prize. In Proceedings of kdd cup and workshop (Vol. 2007, p. 35).
  22. Breese, J. S., Heckerman, D., & Kadie, C. (1998). Empirical analysis of predictive algorithms for collaborative filtering. In Proceedings of the fourteenth conference on uncertainty in artificial intelligence (pp. 43–52).
  23. Friedlander, J. P. (2015). News and notes on 2015 riaa music industry shipment and revenue statistics. Recording Industry Association of America.
  24. Hijikata, Y., Iwahama, K., & Nishida, S. (2006). Content-based music filtering system with editable user profile. In Proceedings of the 2006 acm symposium on applied computing (pp. 1050–1057).
  25. Lew, D. (2007a, Apr). Memory-based algorithms. http://www.cs.carleton.edu/cs comps/0607/recommend/recommender/memorybased.html. (Online)
  26. Lew, D. (2007b, Apr). Model-based recommendation systems. http://www.cs.carleton.edu/cs comps/0607/recommend/recommender/modelbased.html.
  27. Mnih, A., & Salakhutdinov, R. (2007). Probabilistic matrix factorization. In Advances in neural information processing systems (pp. 1257–1264).
  28. Resnick, P., Iacovou, N., Suchak, M., Bergstrom, P., & Riedl, J. (1994). Grouplens: an open architecture for collaborative filtering of netnews. In Proceedings of the 1994 acm conference on computer supported cooperative work (pp. 175–186).
  29. Rubner, Y., Tomasi, C., & Guibas, L. J. (2000). The earth mover’s distance as a metric for image retrieval. International journal of computer vision, 40(2), 99–121.
  30. Van den Oord, A., Dieleman, S., & Schrauwen, B. (2013). Deep content-based music recommendation. In Advances in neural information processing systems (pp. 2643–2651).