中国古典诗歌是一种语言凝练、语义丰富的文学艺术,它的创作因素有许多方面,修辞手法是其中一个最显著的特征之一,诗人在进行创作时通常会使用修辞手法来增强诗歌的感染力和表现力。该文致力于构建具有修辞手法创作能力的诗歌生成模型,以此来提升生成诗歌的多样性、趣味性和新颖性,从而增加读者阅读过程中的审美体验。该文首先通过人工标注、词句特征提取、训练基于BERT的修辞分类器的方式构建一个修辞诗句数据库,然后将每首诗按照一定的方式序列化成一个长句子,并以此来训练语言模型得到诗歌生成模型。自动评测和人工评测结果表明,模型可以生成具有特定修辞手法的诗歌,且生成诗歌的质量相比基线有显著提升。
Abstract
Chinese classical poetry is a kind of literary art with concise language and rich semantics. Its creation involve many aspects, among which rhetoric is one of the most noticeable features. Poets usually use rhetoric to enhance the poem's appeal and expressiveness. This paper aims to build a poetry generation model with the ability of rhetorical creation to improve the diversity, interest, and novelty of generated poems, thus increasing the aesthetic experience of readers in the reading process. To this end, a data set of rhetoric poems is firstly constructed by manual annotation, feature extraction, and a BERT-based rhetoric classifier. Then, each poem is serialized into a sequence, and the poetry generation model is trained as a language model on this sequence. The automatic and manual evaluation results show that the model can generate poems with specific rhetoric, and the quality of the generated poems is significantly improved compared to the baseline.
关键词
中国古典诗歌 /
诗歌生成 /
修辞可控
{{custom_keyword}} /
Key words
Chinese classical poetry /
poem generation /
rhetoric-controllable
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] ZHANG X,LAPATA M. Chinese poetry generation with recurrent neural networks[C]//Proceedings of the Conference on Empirical Methods in Natural Language Processing, 2014: 670-680.
[2] WANG Z, HE W, WU H, et al. Chinese poetry generation with planning based neural network[J].arXiv preprint arXiv: 1610.09889, 2016.
[3] YAN R. Poet: Automatic poetry composition through recurrent neural networks with iterative polishing schema[C]//Proceedings of the IJCAI, 2016: 2238-2244.
[4] YI X, SUN M, LI R, et al. Chinese poetry generation with a working memory model[J].arXiv preprint arXiv: 1809.04306, 2018.
[5] LI J, SONG Y, ZHANG H, et al. Generating classical Chinese poems via conditional variational autoencoder and adversarial training[C]//Proceedings of the Conference on Empirical Methods in Natural Language Processing, 2018: 3890-3900.
[6] 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, 34(05): 9450-9457.
[7] YANG C, SUN M, YI X, et al. Stylistic Chinese poetry generation via unsupervised style disentanglement[C]//Proceedings of the Conference on Empirical Methods in Natural Language Processing, 2018: 3960-3969.
[8] CHEN H, YI X, SUN M, et al. Sentiment-controllable chinese poetry generation[C]//Proceedings of the IJCAI, 2019: 4925-4931.
[9] 周昌乐, 游维, 丁晓君. 一种宋词自动生成的遗传算法及其机器实现[J]. 软件学报, 2010, 21(3): 427-437.
[10] 蒋锐滢, 崔磊, 何晶, 等. 基于主题模型和统计机器翻译方法的中文格律诗自动生成[J]. 计算机学报, 2015, 38(12): 2426-2436.
[11] HE J, ZHOU M, JIANG L. Generating Chinese classical poems with statistical machine translation models[C]//Proceedings of the 26th AAAI Conference on Artificial Intelligence, 2012.
[12] 何晶, 周明, 蒋龙. 基于统计的汉语格律诗生成研究[J]. 中文信息学报, 2010, 24(2): 96-104.
[13] 肖桐, 朱靖波.机器翻译: 基础与模型[M]. 电子工业出版社, 2021.
[14] LUO Y, LI C, HUANG C, et al. Chinese poetry generation with metrical constraints[C]//Proceedings of the CCF International Conference on Natural Language Processing and Chinese Computing. Springer, Cham, 2021: 377-388.
[15] 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.
[16] LIAO Y, WANG Y, LIU Q, et al.Gpt-based generation for classical chinese poetry[J].arXiv preprint arXiv: 1907.00151, 2019.
[17] ZHIPENG G, YI X, SUN M, et al. Jiuge: A human-machine collaborative Chinese classical poetry generation system[C]//Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics: System Demonstrations, 2019: 25-30.
[18] VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems, 2017: 6000-6010.
[19] PAPINENI K, ROUKOS S, WARD T, et al. Bleu: A method for automatic evaluation of machine translation[C]//Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics, 2002: 311-318.
[20] LI J, GALLEY M, BROCKETT C, et al. A diversity-promoting objective function for neural conversation models[J]. arXiv preprint arXiv: 1510.03055, 2015.
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
基金
国家自然科学基金(61876035,61732005)
{{custom_fund}}
PDF(1134 KB)
