吉他和弦譜提供了人們演奏音樂的資訊。網路上有許多吉他和弦譜，但吉他譜的格式不一致，且一首歌通常有許多不同的和弦譜。此外，許多和弦譜的品質不佳，顯示出給定歌曲的和弦序列之間的品質差異。雖然大部分網站提供使用者評等來衡量和弦譜的品質，但新歌曲或不流行歌曲的使用者評等稀少。本篇論文提出一個自動檢驗出和弦譜品質的方法。我們提出深度學習模型學習歌曲和弦序列與和弦譜和弦序列的相似度作為區分和弦譜品質的指標。我們利用此指標對於1000首歌曲與3510張吉他譜做關聯性分析。在實驗中發現，將每張和弦譜轉為與其對應歌曲一致的調後，和弦譜的瀏覽人數與相似度指標有正相關(大於0.12)。相較於其他以編輯距離計算和弦相似度的方法，我們提出的機器學習方法表現較好並且加入音樂特徵以有效地衡量歌曲與吉他譜的相似程度。

Chord tabs provide the information for us to play music. There are many chord tabs for guitar on the Internet. However, they differ in their format, and a song may have many chord tabs. Besides, many chord tabs are of poor quality, and they exhibit the quality discrepancy among chord sequences of a given song. Despite the fact that most websites provide the user ratings, which can be used to measure the quality of chord tabs, user ratings are rare for those unpopular songs and new songs. In this research, we proposed an approach to automatically determine the quality of chord tabs. We propose a deep learning model to learn the chord sequence similarity between a chord tab and its pertaining song, and utilize the similarity as an index to distinguish the quality of chord tabs. We utilize the similarity index to perform the relevance analysis on 1000 songs and 3510 chord tabs. In our experiment, we find that there is a positive correlation between the number of visitors for chord tabs and similarity (above 0.12) after transforming the key of each chord tab to the key of the corresponding songs. Comparing with the other methods which simply calculate chord sequence similarity using editing distance, our proposed machine learning approach performs better and add music features to measure the similarity between songs and chord tabs effectively.

論文審定書 i
誌謝 ii
摘要 iii
Abstract iv
Chapter 1 – Introduction 1
1.1 Background and Motivation 1
1.2 Research Purpose 2
1.3 Thesis Organization 3
Chapter 2-Related Work 4
2.1Chord and Chord Sequence 4
2.2 Chord Sequence Similarity Measurement 7
2.3 Chord Embedding 10
Chapter 3 The Approach 11
3.1 Datasets 13
3.2.1 Ultimate Guitar Chord Tabs 13
3.2.2 RNN Automatic Chord Estimation 13
3.3 Preprocessing of Chords and Chord Sequence 15
3.3.1 Chord Cleaning 15
3.3.2 Key Transfer 16
3.3.3 Chord Embedding 17
3.3.4 Training Data Preparation 18
3.4 Deep Learning Approach 18
3.4.1 Model Architecture 19
3.4.2 Training Process 20
Chapter 4 Experiment 21
4.1 Experiment Dataset 21
4.2 Correlation Test 23
4.3 Ranking Accuracy 25
Chapter 5 Conclusion 29
References 31

1. Berenzweig, A., Logan, B., Ellis, D. P. W., & Whitman, B. (2003). A large-scale evalutation of acoustic and subjective music similarity measures. Proceedings of International Symposium on Music Information Retrieval, 33(2002), 99–105.
2. Berndt, Donald J., Clifford, James (1994) Using Dynamic Time Warping to Find Patterns in Time Series. AAAI Technical Report WS-94-03.
3. Boulanger-Lewandowski, Nicolas,Yoshua, Bengio & Vincent, Pascal (2013) Audio Chord Recognition with Recurrent Neural Networks. ISMIR p. 335-340
4. Casey, M. & Slaney, M. (2006) The Importance of Sequences in Musical Similarity. 2006 IEEE International Conference on Acoustics Speech and Signal Processing Proceedings. Vol. 5.
5. Chai, Wei, Vercoe, Barry (2005). Detection of Key Change in Classical Piano Music. ISMIR, 2005 - ismir2005.ismir.net
6. Cheng, H. T., Yang, Y. H., Lin, Y. C., Liao, I. Bin, & Chen, H. H. (2008). Automatic chord recognition for music classification and retrieval. 2008 IEEE International Conference on Multimedia and Expo, ICME 2008 - Proceedings, 1505–1508. http://doi.org/10.1109/ICME.2008.4607732
7. Haas, W. De. (2008). Tonal pitch step distance: A similarity measure for chord progressions. Proceedings of the International Conference on Music Information Retrieval (ISMIR), (February 2016), 51–56.
8. Hanna, P., Ferraro, P., Robine, M., Hanna, P., Ferraro, P., Robine, M.,Robine, M. (2007). On Optimizing the Editing Algorithms for Evaluating Similarity Between Monophonic Musical Sequences. Journal of New Music Research, 36, 267–279.
9. Hrybyk Alex and Kim Youngmoo E. (2010). Combined audio and video analysis for guitar chord identification. In Proc 11th ISMIR, pages 159-164.
10. Huang, C.-Z. A., Duvenaud, D., and Gajos, K. Z. (2016) Chordripple: Recommending chords to help novice composers go beyond the ordinary. In Proceedings of the Conference on Intelligent User Interfaces.
11. Katayose, H., Imai, M., & Inokuchi, S. (1988). Sentiment extraction in music. In: [1988 Proceedings] 9th International Conference on Pattern Recognition. IEEE, 1988. p. 1083-1087.
12. Levenshtein, V. I. (1966) Binary Codes Capable of Correcting deletions, insertions, and reversals. Soviet Physics, 10(8): 707-710.
13. Macrae, R., & Dixon, S. (2011). Guitar tab mining, analysis and ranking. 12th International Society for Music Information Retrieval Conference (ISMIR 2011), (Ismir), 453–458.
14. Madjiheurem, Sephora, Qu, Lizhen & Walder, Christian: (2016) Chord2Vec: Learning Musical Chord Embeddings. In: Proceedings of the constructive machine learning workshop at 30th conference on neural information processing systems (NIPS’2016), Barcelona, Spain. 2016. p. 1-5.
15. Mauch, M., & Dixon, S. (2010). Simultaneous estimation of chords and musical context from audio. IEEE Transactions on Audio, Speech and Language Processing, 18(6), 1280–1289. http://doi.org/10.1109/TASL.2009.2032947
16. McVicar, Matt, Santos-Rodríguez, Raúl, Ni Yizhao, and Bie, Tijl De (2014) Automatic Chord Estimation from Audio: A Review of the State of the Art. IEEE/ACM Transactions on Audio, Speech, and Language Processing, VOL. 22, NO. 2
17. Mikolov, Tomas, Chen, Kai, Corrado, Greg, and Dean Jeffrey. (2013).Efficient estimation of word representations in vector space. CoRR, abs/1301.3781, 2013
18. Randel, Don Michael. (1999). The Harvard concise dictionary of music and musicians. Harvard University Press.
19. Wongsaroj, C., Prompoon, N., & Surarerks, A. (2014). A music similarity measure based on chord progression and song segmentation analysis. 2014 Fourth International Conference on Digital Information and Communication Technology and Its Applications (DICTAP), 158–163.
20. Zhou, Xinguan, Lerch, Alexander (2015) Chord Detection Using Deep Learning. IEEE International Conference on Acoustics, Speech and Signal Processing(ICASSP)