請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/78795
完整後設資料紀錄
DC 欄位 | 值 | 語言 |
---|---|---|
dc.contributor.advisor | 陳宏銘(Homer H. Chen) | |
dc.contributor.author | Yi-Hsuan Wu | en |
dc.contributor.author | 吳奕萱 | zh_TW |
dc.date.accessioned | 2021-07-11T15:19:54Z | - |
dc.date.available | 2022-12-31 | |
dc.date.copyright | 2020-09-23 | |
dc.date.issued | 2020 | |
dc.date.submitted | 2020-08-29 | |
dc.identifier.citation | [1] M. Fell, E. Cabrio, F. Gandon, and A. Giboin, “Song lyrics summarization inspired by audio thumbnailing,” in Recent Advances in Natural Language Processing, 2019. [2] T. Zhang, V. Kishore, F. Wu, K. Q. Weinberger, and Y. Artzi, “Bertscore: Evaluating text generation with bert,” in International Conference on Learning Representations (ICLR), 2020. [3] R. Hossain, Md. R. K. R. Sarker, M. Mimo, A. A. Marouf, and B. Pandey, “Recommendation approach of English songs title based on Latent Dirichlet Allocation applied on lyrics,” In 2019 IEEE International Conference on Electrical, Computer and Communication Technologies (ICECCT), IEEE, 2019. p. 1-4. [4] R. Mihalcea and P. Tarau, “TextRank: Bringing order into texts,” in Proc. Conf. Empirical Methods Natural Language Processing, 2004. [5] Q. Zhou, N. Yang, F. Wei, S. Huang, M. Zhou, and T. Zhao, “Neural document summarization by jointly learning to score and select sentences,” in Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (ACL), pp. 654–663, 2018. [6] R. Nallapati, F. Zhai, and B. Zhou, “SummaRuNNer: A recurrent neural network based sequence model for extractive summarization of documents,” in Proceedings of the Thirty-First AAAI Conference on Artificial Intelligence (AAAI-17), 2017. [7] Abigail See, Peter J. Liu, and Christopher D. Manning, “Get to the point: Summarization with pointer-generator networks,” in Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pp. 1073–1083, 2017. [8] A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, and I. Polosukhin, “Attention is all you need,” In Advances in neural information processing systems (NIPS), pp. 5998–6008, 2017. [9] M. Lewis, Y. Liu, N. Goyal, M. Ghazvininejad, A. Mohamed, O. Levy, V. Stoyanov, and L. Zettlemoyer, “BART: Denoising sequence-to-sequence pre-training for natural language generation, translation, and comprehension,” CoRR, abs/1910.13461, 2019. [10] Y. Liu and M. Lapata, “Text summarization with pretrained encoders,” in Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing (EMNLP), pp. 3730–3740, 2019. [11] N. S. Keskar, B. McCann, L. R. Varshney, C. Xiong, and R. Socher, “Ctrl: A conditional transformer language model for controllable generation,” ArXiv, abs/1909.05858. [12] J. Devlin, M.-W. Chang, K. Lee, and K. Toutanova, “BERT: Pre-training of deep bidirectional transformers for language understanding,” in Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pp. 4171–4186, 2019. [13] S. Rothe, S. Narayan, and A. Severyn, “Leveraging Pre-trained Checkpoints for Sequence Generation Tasks,” Transactions of the Association for Computational Linguistics 2020 Vol. 8, pp. 264-280 [14] A. Radford, J. Wu, R. Child, D. Luan, D. Amodei, and I. Sutskever, “Language models are unsupervised multitask learners,” Technical report, OpenAI. [15] Y. Liu, M. Ott, N. Goyal, J. Du, M. Joshi, D. Chen, O. Levy, M. Lewis, L. Zettlemoyer, and V. Stoyanov, “RoBERTa: A robustly optimized BERT pretraining approach,” arxiv preprint arXiv:1907.11692. [16] J. Zhang, Y. Zhao, M. Saleh, and P. J. Liu, “PEGASUS: Pre-training with extracted gap-sentences for abstractive summarization,” in Proceedings of the 37th International Conference on Machine Learning, PMLR 108, 2020. [17] J. Wieting, T. Berg-Kirkpatrick, K. Gimpel, and G. Neubig, “Beyond BLEU: Training neural machine translation with semantic similarity,” in Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics (ACL), pp. 4344–4355, 2019. [18] J.-P. Ng and V. Abrecht, “Better summarization evaluation with word embeddings for ROUGE,” CoRR, abs/1508.06034, 2015. [19] W. Kryscinski, N. S. Keskar, B. McCann, C. Xiong, and R. Socher, “Neural text summarization: A critical evaluation,” in Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing, pp. 540–551, 2019. [20] H. Zheng and M. Lapata, “Sentence centrality revisited for unsupervised summarization,” in Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics (ACL), pp. 6236–6247, 2019. [21] S.E. Robertson, “Understanding inverse document frequency: On theoretical arguments for IDF,” Journal of Documentation, vol. 60, no. 5, pp. 503-520, 2004 [22] N. Reimers and I. Gurevych, “Sentence-BERT: Sentence embeddings using Siamese BERT-networks,” in Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics (ACL), pp. 3982–3992, 2019. [23] C. Raffel, N. Shazeer, A. Roberts, K. Lee, S. Narang, M. Matena, Y. Zhou, W. Li and P. J. Liu, “Exploring the limits of transfer learning with a unified text-to-text transformer,” in Journal of Machine Learning Research, vol. 21, pp. 1-67, 2020. | |
dc.identifier.uri | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/78795 | - |
dc.description.abstract | 歌曲賞析能讓使用者快速掌握一首歌曲的情境內容,判別這首歌是否符合當下的聆聽需求。歌曲賞析可以利用文字摘要模型從歌詞生成,然而歌詞的抽象特性使得現今文字摘要模型往往無法掌握其涵義,而產生與內容不相關的語句。在這篇論文裡,我們藉由資料處理方法之設計與訓練目標函數之設定,以產生符合歌曲主題的文字敘述。具體而言,我們從網路論壇蒐集歌曲評論作為資料集,並且將歌詞精華與之結合。歌詞精華屬於萃取式(extractive)摘要,是利用語句關係網路以及句子在向量空間的分布,由歌詞選取而得;歌曲評論屬於概括式(abstractive)摘要,來自網路論壇。我們的訓練資料是以萃取式和概括式兩種形式的摘要組合而成,以避免受到歌曲評論的雜論度影響。我們的模型架構為基於注意力機制之神經網路(Transformer),以成對的歌詞與摘要作為訓練素材,學習產生概括式摘要。模型的目標函數結合最大似然估計(maximum likelihood estimation)與向量相似度。實驗結果顯示,此研究所探討的資料處理方法與訓練目標函數有助於提升文字摘要模型的表現。 | zh_TW |
dc.description.abstract | Music descriptions help users understand the context of a song at a glance. Given the figurative nature of song lyrics, current text summarization models often fail to capture the meaning expressed in songs and, as a result, generate imaginative but irrelevant descriptions. In this work, we propose a music description (or summary) generation scheme based on a novel data representation and training objective. The generation of music descriptions is built upon a Transformer-based model, for which the training objective incorporates semantic similarity into maximum likelihood estimation (MLE). To combat noise, our reference summary for the data representation of a song contains both extractive and abstractive components obtained from the lyrics highlight and interpretation of the song. The lyrics highlight is obtained from graph-based ranking and embedding similarity. The data representation serves as the pseudo-ground-truth for sequence-to-sequence abstractive summarization. The effectiveness of the purposed method is evaluated by metrics such as ROUGE, BLEU, and BERTScore. | en |
dc.description.provenance | Made available in DSpace on 2021-07-11T15:19:54Z (GMT). No. of bitstreams: 1 U0001-1908202016173400.pdf: 1782228 bytes, checksum: e59940b6003f73aded36f3380b32c281 (MD5) Previous issue date: 2020 | en |
dc.description.tableofcontents | Chapter 1 Introduction………………………………………………………………………1 Chapter 2 Related Work………………………………………………………………………4 2.1 Music Descriptions………………………………………………………4 2.2 Text Summarization………………………………………………………4 2.3 Semantic Similarity……………………………………………………6 2.4 Evaluation Criteria and Methods……………………6 Chapter 3 Proposed Lyrics Summarization Method………8 3.1 Data Representation and Preprocessing…………8 3.2 Model……………………………………………………………………………………………10 3.3 Training Objective…………………………………………………………12 Chapter 4 Experiments………………………………………………………………………15 4.1 Dataset………………………………………………………………………………………15 4.2 Data Preprocessing…………………………………………………………17 4.2.1 Rating-based criteria………………………17 4.2.2 Quote-based criterion………………………17 4.2.3 Label-based criteria…………………………18 4.3 Constructing Training Samples……………………………19 4.4 Experimental Setup…………………………………………………………20 4.5 Performance Evaluation.……………………………………………21 4.6 Evaluation Metrics.………………………………………………………22 Chapter 5 Results and Discussion…………………………………………23 5.1 Efficacy of Quote-enhanced Training Samples……………23 5.2 Efficacy of Data Preprocessing………………………………………………25 5.3 Efficacy of Objectives Functions…………………………………………28 5.4 Discussions…………………………………………………………………………………………………33 Chapter 6 Conclusion…………………………………………………………………………36 | |
dc.language.iso | en | |
dc.title | 從歌詞產生歌曲賞析 | zh_TW |
dc.title | Lyrics Transformer: Generating Music Descriptions from Pop Song Lyrics | en |
dc.type | Thesis | |
dc.date.schoolyear | 108-2 | |
dc.description.degree | 碩士 | |
dc.contributor.oralexamcommittee | 蔡銘峰(Ming-Feng Tsai),楊奕軒(Yi-Hsuan Yang),王家慶(Jia-Ching Wang),李宏毅(Hung-Yi Lee) | |
dc.subject.keyword | 文章摘要,歌詞,歌曲賞析,音樂周邊資訊,音樂資訊檢索, | zh_TW |
dc.subject.keyword | text summarization,song lyrics,music description,music contextual information,music information retrieval, | en |
dc.relation.page | 39 | |
dc.identifier.doi | 10.6342/NTU202004092 | |
dc.rights.note | 有償授權 | |
dc.date.accepted | 2020-08-31 | |
dc.contributor.author-college | 電機資訊學院 | zh_TW |
dc.contributor.author-dept | 電信工程學研究所 | zh_TW |
顯示於系所單位: | 電信工程學研究所 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
U0001-1908202016173400.pdf 目前未授權公開取用 | 1.74 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。