融合韵律特征的诗歌生成模型

doi:10.3778/j.issn.1002-8331.2303-0477

摘要/Abstract

摘要： 诗歌生成中的韵律规范和主题一致性一直以来都是自然语言生成领域的研究热点。为提升诗歌生成中的韵律规范，提出了基于Transformer结合韵律特征的诗歌生成模型（Transformer and prosodic features poetry generation model，TPPG）。根据韵律特征建立平仄韵律词库和平声韵脚词库，在Transformer编码器中引入平仄韵律编码，模型训练过程中可以捕获更多平仄韵律特征的信息，学习到多种诗歌韵律；最终根据建立的平声韵脚词库规范诗歌生成韵脚，运用极大后验概率对于候选的诗歌选择当前赋有韵律特征规范的最优诗句，整体提升诗歌规范性和流畅性。实验结果表明TPPG模型生成的诗歌能够很好地符合韵律，在人工评价和机器评价中均有提高。

关键词: 诗歌生成, 韵律库, 韵律编码, 韵律特征

Abstract: Prosody specification and topic consistency in poetry generation have always been a research hotspot in the field of natural language generation. In order to improve the prosody specification in poetry generation, this paper proposes a poem generation model based on Transformer combined with prosody (Transformer and prosodic features poetry generation model, TPPG). According to the prosodic features, the prosodic thesaurus and the prosodic thesaurus are established, and the prosodic coding is introduced into the Transformer encoder. During the model training process, more prosodic feature information can be captured, and a variety of poetic rhythms can be learned. Finally, the rhyme is generated according to the established rhyme lexicon, and the optimal verse with the rhythm feature specification is selected for the candidate poem by using a large posterior probability, which improves the standardization and fluency of the poem as a whole. The experimental results show that the poetry generated by TPPG model can conform to rhythm well, and it has improved in both manual evaluation and machine evaluation.

Key words: poetry generation, rhythm library, prosodic encoding, prosodic features

吴林东, 何向真, 万福成. 融合韵律特征的诗歌生成模型[J]. 计算机工程与应用, 2024, 60(13): 162-170.

WU Lindong, HE Xiangzhen, WAN Fucheng. Poetry Generative Model Incorporating Prosodic Features[J]. Computer Engineering and Applications, 2024, 60(13): 162-170.

参考文献

[1] GHAZVININEJAD M, SHI X, CHOI Y, et al. Generating topical poetry[C]//Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, 2016: 1183-1191.
[2] YI X, SUN M, LI R, et al. Chinese poetry generation with a working memory model[C]//Proceedings of the International Joint Conference on Artificial Intelligence, 2018: 4553-4559.
[3] 廖荣凡, 沈希忠, 刘爽. 更具有感情色彩的诗歌生成模型[J]. 计算机系统应用, 2020, 29(5): 46-51.
LIAO R F, SHEN X Z, LIU S. Poetry generation model with more emotional information[J]. Computer Systems Applications, 2020, 29(5): 46-51.
[4] 叶旺. 格式限制的古诗可控生成学习研究[D]. 杭州: 杭州电子科技大学, 2022.
YE W. Towards conditional text generation in formats[D]. Hangzhou: Hangzhou Dianzi University, 2022.
[5] YI X, LI R, YANG C, et al. Mixpoet: diverse poetry generation via learning controllable mixed latent space[C]//Proceedings of the AAAI Conference on Artificial Intelligence, 2020: 9450-9457.
[6] 王勇超, 周灵智, 赵亚萍, 等. 自动扩充关键词语义信息的诗歌生成算法[J]. 计算机科学与探索, 2023, 17(6): 1387-1394.
WANG Y C, ZHOU L Z, ZHAO Y P, et al. Poetry generation algorithm with automatic expansion of keyword semantic information[J]. Journal of Frontiers of Computer Science and Technology, 2023, 17(6): 1387-1394.
[7] LIU Z, FU Z, CAO J, et al. Rhetorically controlled encoder decoder for modern Chinese poetry generation[C]//Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, 2019: 1992-2001.
[8] 李章斌. “韵”之离散: 关于当代中国诗歌韵律的一种观察[J]. 中国当代文学研究, 2020(3): 52-66.
LI Z B. The discreteness of “rhyme”: an observation on the rhythm of contemporary Chinese poetry[J]. Contemporary Chinese Literature Studies, 2020(3): 52-66.
[9] YI X, LI R, SUN M. Generating Chinese classical poems with RNN encoder-decoder[C]//Proceedings of the China National Conference on Chinese Computational Linguistics International Symposium on Natural Language Processing Based on Naturally Annotated Big Data, 2017: 211-223.
[10] YANG C, SUN M, YI X, et al. Stylistic Chinese poetry generation via unsupervised style disentanglement[C]//Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, 2018: 3960-3969.
[11] H?M?L?INEN M, ALNAJJAR K, POIBEAU T. Modern French poetry generation with RoBERTa and GPT-2[C]//Proceedings of the 13th International Conference on Computational Creativity, 2022: 12-16.
[12] VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C]//Advances in Neural Information Processing Systems, 2017, 30: 5998-6008.
[13] LIU Y, LIU D, LV J, et al. Generating Chinese poetry from images via concrete and abstract information[C]//Proceedings of the 2020 International Joint Conference on Neural Networks, 2020.
[14] WANG J, LI H, WU C, et al. Generating diverse Chinese poetry from images via unsupervised method[J]. Neurocomputing, 2022, 492: 188-200.
[15] BAHDANAU D, CHO K, BENGIO Y. Neural machine translation by jointly learning to align and translate[C]//Proceedings of the ICLR, 2015: 1-15.
[16] WANG D Z, HE W, WU H, et al. Chinese poetry generation with planning based neural network[C]//Proceedings of the 26th International Conference on Computational Linguistics: Technical Papers, 2016: 1051-1060.
[17] ZHANG X, LAPATA M. Chinese poetry generation with recurrent neural networks[C]//Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing, 2014: 670-680.
[18] YANG X, LIN X, SUO S, et al. Generating thematic Chinese poetry using conditional variational autoencoders with hybrid decoders[C]//Proceedings of the 27th International Joint Conference on Artificial Intelligence, 2018: 4539-4545.