We use cookies to improve your experience with our site.

Indexed in:

SCIE, EI, Scopus, INSPEC, DBLP, CSCD, etc.

Submission System
(Author / Reviewer / Editor)
He M, Chen Y, Zhao HK et al. Composing like an ancient Chinese poet: Learn to generate rhythmic Chinese poetry. JOURNAL OF COMPUTER SCIENCE AND TECHNOLOGY 38(6): 1272−1287 Nov. 2023. DOI: 10.1007/s11390-023-1295-1.
Citation: He M, Chen Y, Zhao HK et al. Composing like an ancient Chinese poet: Learn to generate rhythmic Chinese poetry. JOURNAL OF COMPUTER SCIENCE AND TECHNOLOGY 38(6): 1272−1287 Nov. 2023. DOI: 10.1007/s11390-023-1295-1.

Composing Like an Ancient Chinese Poet: Learn to Generate Rhythmic Chinese Poetry

Funds: This work was supported by the National Natural Science Foundation of China under Grant Nos. 61922073 and 72101176.
More Information
  • Author Bio:

    Ming He received his Ph.D. degree in computer science from University of Science and Technology of China, Hefei, in 2018. He is currently an advisory researcher with the AI Laboratory, Lenovo Research, Beijing. He was a visiting scholar at University of North Carolina at Charlotte in 2016–2017, was an algorithm researcher of DiDi, TAL (Tomorrow Advancing Life) in 2019–2021, and finished his post-doctor program at Department of Electronic Engineering of Shanghai Jiao Tong University, Shanghai, in 2022. His current research interests include recommendation system and reinforcement learning. He has published more than 15 papers in refereed journals and conference proceedings, e.g., TIP and SIGIR. He received the KSEM 2018 Best Research Paper Award and submitted more than 120 patents to Chinese National Patent Office

    Yan Chen received her M.S. degree in computer science from University of Science and Technology of China, Hefei, in 2018. She is currently a senior engineer with Baidu Incorporated, Beijing. Her current research interests include model robust, recommendation system and question answering system. She has published more than five papers in refereed journals and conference proceedings, e.g., EMNLP, DASFAA, and KSEM. She received the KSEM 2018 Best Research Paper Award

    Hong-Ke Zhao received his Ph.D. degree in computer science from University of Science and Technology of China, Hefei, in 2019. He is currently an associate professor with the College of Management and Economics, Tianjin University, Tianjin. His research interests include data mining, data-driven management, and knowledge and behavior computing. He has published more than 70 papers in refereed journals and conference proceedings, such as INFORMS Journal on Computing, TKDE, TBD, Industrial Marketing Management, Scientometrics, KDD, WWW, IJCAI and AAAI. He was the recipient of Distinguished Dissertation Award Nomination of CAAI (2019), and the Best Student Paper Award of CCML 2019. He has served regularly on the program committees of conferences, including KDD, AAAI, IJCAI, and WWW

    Qi Liu received his Ph.D. degree in computer science from University of Science and Technology of China, Hefei, in 2013. He is currently a professor with the School of Computer Science and Technology, University of Science and Technology of China, Hefei. His research interests include data mining, machine learning, and recommender systems. He was the recipient of KDD 2018 Best Student Paper Award and ICDM 2011 Best Research Paper Award. He was also the recipient of China Outstanding Youth Science Foundation in 2019. He is a member of IEEE

    Le Wu received her Ph.D. degree in computer science from University of Science and Technology of China, Hefei, in 2015. She is an associate professor with the School of Computer Science and Information Engineering, Hefei University of Technology, Hefei. Her general area of research is data mining, recommendation system, and social network analysis. She has published several papers in referred journals and conferences, such as the IEEE Transactions on Knowledge and Data Engineering, the ACM Transactions on Intelligent Systems and Technology, AAAI, IJCAI, KDD, SDM, and ICDM. She is the recipient of the Best of SDM 2015 Award. She is a member of IEEE

    Yu Cui received her M.E. degree in business administration from Beijing Institute of Technology, Beijing, in 2021. She is currently an algorithm engineer with the Risk Control Department, Opay, Beijing. She was an NLP algorithm engineer of TAL (Tomorrow Advancing Life) from 2018 to 2021. Her current research interests include text mining, dialogue systems, sentiment analysis, content risk control, and machine learning

    Gui-Hua Zeng received his Ph.D. degree in optics from Shanghai Institute of Optics and Fine Mechanics, Chinese Academy of Sciences, Shanghai, in 1997. He is currently a professor with the Department of Electronic Engineering, Shanghai Jiao Tong University, Shanghai. His research interests include quantum cryptography, quantum optical communication, and quantum machine learning. He has published over 200 papers in refereed conferences and journals, e.g., IEEE Trans., Phys. Rev. A, and NPJ Quantum Information

    Gui-Quan Liu received his M.S. and Ph.D. degrees in computer science from University of Science and Technology of China, Hefei, in 1996 and 1999, respectively. He is currently an associate professor with the School of Computer Science and Technology, University of Science and Technology of China, Hefei. His main research interests include machine learning, data mining, social network analysis, and pattern recommendation. He has published over 60 papers in refereed conferences and journals

  • Corresponding author:

    ghzeng@sjtu.edu.cn

  • Co-First Author (Ming He and Yan Chen contributed equally to this work.)

  • Received Date: January 15, 2021
  • Accepted Date: September 17, 2023
  • Automatic generation of Chinese classical poetry is still a challenging problem in artificial intelligence. Recently, Encoder-Decoder models have provided a few viable methods for poetry generation. However, by reviewing the prior methods, two major issues still need to be settled: 1) most of them are one-stage generation methods without further polishing; 2) they rarely take into consideration the restrictions of poetry, such as tone and rhyme. Intuitively, some ancient Chinese poets tended first to write a coarse poem underlying aesthetics and then deliberated its semantics; while others first create a semantic poem and then refine its aesthetics. On this basis, in order to better imitate the human creation procedure of poems, we propose a two-stage method (i.e., restricted polishing generation method) of which each stage focuses on the different aspects of poems (i.e., semantics and aesthetics), which can produce a higher quality of generated poems. In this way, the two-stage method develops into two symmetrical generation methods, the aesthetics-to-semantics method and the semantics-to-aesthetics method. In particular, we design a sampling method and a gate to formulate the tone and rhyme restrictions, which can further improve the rhythm of the generated poems. Experimental results demonstrate the superiority of our proposed two-stage method in both automatic evaluation metrics and human evaluation metrics compared with baselines, especially in yielding consistent improvements in tone and rhyme.

  • [1]
    Hopkins J, Kiela D. Automatically generating rhythmic verse with neural networks. In Proc. the 55th Annual Meeting of the Association for Computational Linguistics, July 2017, pp.168–178. DOI: 10.18653/v1/P17-1016.
    [2]
    He J, Zhou M, Jiang L. Generating Chinese classical poems with statistical machine translation models. In Proc. the 26th AAAI Conference on Artificial Intelligence, July 2012, pp.1650–1656. DOI: 10.1609/aaai.v26i1.8344.
    [3]
    Wang Z, He W, Wu H, Wu H Y, Li W, Wang H F, Chen E H. Chinese poetry generation with planning based neural network. arXiv: 1610.09889, 2016. https://arxiv.org/abs/1610.09889, Sept. 2023.
    [4]
    Oliveira H G, Cardoso A. Poetry generation with PoeTryMe. In Computational Creativity Research: Towards Creative Machines, Besold T R, Schorlemmer M, Smaill A (eds.), Atlantis Press, 2015, pp.243–266. DOI: 10.2991/978-94-6239-085-0_12.
    [5]
    Tosa N, Obara H, Minoh M. Hitch Haiku: An interactive supporting system for composing haiku poem. In Proc. the 7th International Conference on Entertainment Computing, Sept. 2008, pp.209–216. DOI: 10.1007/978-3-540-89222-9_26.
    [6]
    Manurung H. An evolutionary algorithm approach to poetry generation [Ph.D. Thesis]. School of Informatics, The University of Edinburgh, Edinburgh, 2004.
    [7]
    Manurung R, Ritchie G, Thompson H. Using genetic algorithms to create meaningful poetic text. Journal of Experimental and Theoretical Artificial Intelligence, 2012, 24(1): 43–64. DOI: 10.1080/0952813X.2010.539029.
    [8]
    Jiang L, Zhou M. Generating Chinese couplets using a statistical MT approach. In Proc. the 22nd International Conference on Computational Linguistics, Aug. 2008, pp.377–384. DOI: 10.3115/1599081.1599129.
    [9]
    Bahdanau D, Cho K, Bengio Y. Neural machine translation by jointly learning to align and translate. arXiv: 1409.0473, 2016. https://arxiv.org/abs/1409.0473, Sept. 2023.
    [10]
    Yi X Y, Li R Y, Sun M S. Generating Chinese classical poems with RNN encoder-decoder. In Proc. the 2017 Conference on Chinese Computational Linguistics and Natural Language Processing Based on Naturally Annotated Big Data, Oct. 2017, pp.211–223. DOI: 10.1007/978-3-319-69005-6_18.
    [11]
    Zhang X X, Lapata M. Chinese poetry generation with recurrent neural networks. In Proc. the 2014 Conference on Empirical Methods in Natural Language Processing, Oct. 2014, pp.670–680. DOI: 10.3115/v1/D14-1074.
    [12]
    Gervás P. An expert system for the composition of formal Spanish poetry. Knowledge-Based Systems, 2001, 14(3/4): 181–188. DOI: 10.1016/S0950-7051(01)00095-8.
    [13]
    Wu X F, Tosa N, Nakatsu R. New hitch haiku: An interactive Renku poem composition supporting tool applied for sightseeing navigation system. In Proc. the 8th International Conference on Entertainment Computing, Sept. 2009, pp.191–196. DOI: 10.1007/978-3-642-04052-8_19.
    [14]
    Yan R, Jiang H, Lapata M, Lin S D, Lv X Q, Li X M. i, Poet: Automatic Chinese poetry composition through a generative summarization framework under constrained optimization. In Proc. the 23rd International Joint Conference on Artificial Intelligence, Aug. 2013, pp.2197–2203. DOI: 10.5555/2540128.2540444.
    [15]
    Jiang N, Chen J, Zhou R G, Wu C X, Chen H L, Zheng J Q, Wan T. PAN: Pipeline assisted neural networks model for data-to-text generation in social Internet of Things. Information Sciences, 2020, 530: 167–179. DOI: 10.1016/j.ins.2020.03.080.
    [16]
    Ameur H, Jamoussi S, Hamadou A B. A new method for sentiment analysis using contextual auto-encoders. Journal of Computer Science and Technology, 2018, 33(6): 1307–1319. DOI: 10.1007/s11390-018-1889-1.
    [17]
    Bengio Y, Ducharme R, Vincent P, Janvin C. A neural probabilistic language model. The Journal of Machine Learning Research, 2003, 3: 1137–1155. DOI: 10.1162/153244303322533223.
    [18]
    Goldberg Y. Neural Network Methods for Natural Language Processing. Morgan and Claypool Publishers, 2017.
    [19]
    Lai Y N, Zhang L F, Han D H, Zhou R, Wang G R. Fine-grained emotion classification of Chinese microblogs based on graph convolution networks. World Wide Web, 2020, 23(5): 2771–2787. DOI: 10.1007/s11280-020-00803-0.
    [20]
    Li Y, Pan Q, Wang S H, Yang T, Cambria E. A generative model for category text generation. Information Sciences, 2018, 450: 301–315. DOI: 10.1016/j.ins.2018.03.050.
    [21]
    Tao H Q, Tong S W, Zhao H K, Xu T, Jin B B, Liu Q. A radical-aware attention-based model for Chinese text classification. In Proc. the 33rd AAAI Conference on Artificial Intelligence, Jan. 2019, pp.5125–5132. DOI: 10.1609/aaai.v33i01.33015125.
    [22]
    Yuan C Z, Bao Z F, Sanderson M, Tang Y. Incorporating word attention with convolutional neural networks for abstractive summarization. World Wide Web, 2020, 23(1): 267–287. DOI: 10.1007/s11280-019-00709-6.
    [23]
    Zhu H Y, Liu Q, Yuan N J, Zhang K, Chen E H. Pop music generation: From melody to multi-style arrangement. ACM Trans. Knowledge Discovery from Data, 2020, 14(5): Article No. 54. DOI: 10.1145/3374915.
    [24]
    Wu J J, Yang W, Li L D, Dong W S, Shi G M, Lin W S. Blind image quality prediction with hierarchical feature aggregation. Information Sciences, 2021, 552: 167–182. DOI: 10.1016/j.ins.2020.12.018.
    [25]
    Jiang T B, Xia G S, Lu Q K, Shen W M. Retrieving aerial scene images with learned deep image-sketch features. Journal of Computer Science and Technology, 2017, 32(4): 726–737. DOI: 10.1007/s11390-017-1754-7.
    [26]
    Liu H, Fu Z L, Han J G, Shao L, Hou S D, Chu Y Z. Single image super-resolution using multi-scale deep encoder-decoder with phase congruency edge map guidance. Information Sciences, 2019, 473: 44–58. DOI: 10.1016/j.ins.2018. 09.018.
    [27]
    Pawar K, Attar V. Deep learning approaches for video-based anomalous activity detection. World Wide Web, 2019, 22(2): 571–601. DOI: 10.1007/s11280-018-0582-1.
    [28]
    Wu Y, Ma Y J, Liu J, Du J, Xing L. Self-attention convolutional neural network for improved MR image reconstruction. Information Sciences, 2019, 490: 317–328. DOI: 10.1016/j.ins.2019.03.080.
    [29]
    Ghazvininejad M, Shi X, Priyadarshi J, Knight K. Hafez: An interactive poetry generation system. In Proc. the ACL 2017, System Demonstrations, July 2017, pp.43–48. DOI: 10.18653/v1/P17-4008.
    [30]
    Wang Q X, Luo T Y, Wang D, Xing C. Chinese song iambics generation with neural attention-based model. In Proc. the 25th International Joint Conference on Artificial Intelligence, July 2016, pp.2943–2949. DOI: 10.5555/3060832.3061033.
    [31]
    Mikolov T, Karafiát M, Burget L, Černocký J, Khudanpur S. Recurrent neural network based language model. In Proc. the 11th Annual Conference of the International Speech Communication Association, Sept. 2010, pp.1045–1048. DOI: 10.21437/Interspeech.2010-343.
    [32]
    Wang Q X, Luo T Y, Wang D. Can machine generate traditional Chinese poetry? A feigenbaum test. In Proc. the 8th International Conference on Brain Inspired Cognitive Systems, Nov. 2016, pp.34–46. DOI: 10.1007/978-3-319-49685-6_4.
    [33]
    Li J T, Song Y, Zhang H S, Chen D M, Shi S M, Zhao D Y, Yan R. Generating classical Chinese poems via conditional variational autoencoder and adversarial training. In Proc. the 2018 Conference on Empirical Methods in Natural Language Processing, Oct. 31–Nov. 4, 2018, pp.3890–3900. DOI: 10.18653/v1/D18-1423.
    [34]
    Yi X Y, Li R Y, Sun M S. Chinese poetry generation with a salient-clue mechanism. In Proc. the 22nd Conference on Computational Natural Language Learning, Oct. 2018, pp.241–250. DOI: 10.18653/v1/K18-1024.
    [35]
    Zhang J Y, Feng Y, Wang D, Wang Y, Abel A, Zhang S Y, Zhang A D. Flexible and creative Chinese poetry generation using neural memory. In Proc. the 55th Annual Meeting of the Association for Computational Linguistics, July 2017, pp.1364–1373. DOI: 10.18653/v1/P17-1125.
    [36]
    Yang X P, Lin X W, Suo S D, Li M. Generating thematic Chinese poetry using conditional variational autoencoders with hybrid decoders. In Proc. the 27th International Joint Conference on Artificial Intelligence, July 2018, pp.4539–4545. DOI: 10.24963/ijcai.2018/631.
    [37]
    Xu L L, Jiang L, Qin C, Wang Z, Du D F. How images inspire poems: Generating classical Chinese poetry from images with memory networks. In Proc. the 32nd AAAI Conference on Artificial Intelligence, Feb. 2018, pp.5618–5625. DOI: 10.1609/aaai.v32i1.12001.
    [38]
    Liu D, Guo Q, Li W B, Lv J C. A multi-modal Chinese poetry generation model. In Proc. the 2018 International Joint Conference on Neural Networks, July 2018. DOI: 10.1109/IJCNN.2018.8489579.
    [39]
    Liu Y S, Liu D, Lv J C. Deep poetry: A Chinese classical poetry generation system. In Proc. the 34th AAAI Conference on Artificial Intelligence, Feb. 2020, pp.13626–13627. DOI: 10.1609/aaai.v34i09.7100.
    [40]
    Ghazvininejad M, Shi X, Choi Y J, Knight K. Generating topical poetry. In Proc. the 2016 Conference on Empirical Methods in Natural Language Processing, Nov. 2016, pp.1183–1191. DOI: 10.18653/v1/D16-1126.
    [41]
    Greene E, Bodrumlu T, Knight K. Automatic analysis of rhythmic poetry with applications to generation and translation. In Proc. the 2010 Conference on Empirical Methods in Natural Language Processing, Oct. 2010, pp.524–533. DOI: 10.5555/1870658.1870709.
    [42]
    Kolesnikov A, Lampert C H. PixelCNN models with auxiliary variables for natural image modeling. arXiv: 1612.08185, 2016. https://arxiv.org/abs/1612.08185, Sept. 2023.
    [43]
    Gu J X, Cai J F, Wang G, Chen T. Stack-captioning: Coarse-to-fine learning for image captioning. In Proc. the 32nd AAAI Conference on Artificial Intelligence, Feb. 2018, pp.6837–6844. DOI: 10.1609/aaai.v32i1.12266.
    [44]
    Guu K, Hashimoto T B, Oren Y, Liang P. Generating sentences by editing prototypes. Trans. Association for Computational Linguistics, 2018, 6: 437–450. DOI: 10.1162/tacl_a_00030.
    [45]
    Pandey G, Contractor D, Kumar V, Joshi S. Exemplar encoder-decoder for neural conversation generation. In Proc. the 56th Annual Meeting of the Association for Computational Linguistics, July 2018, pp.1329–1338. DOI: 10.18653/v1/P18-1123.
    [46]
    Wang Q Y, Zhou Z H, Huang L F, Whitehead S, Zhang B L, Ji H, Knight K. Paper abstract writing through editing mechanism. In Proc. the 56th Annual Meeting of the Association for Computational Linguistics, July 2018, pp.260–265. DOI: 10.18653/v1/P18-2042.
    [47]
    Xia Y C, Tian F, Wu L J, Lin J X, Qin T, Yu N H, Liu T Y. Deliberation networks: Sequence generation beyond one-pass decoding. In Proc. the 31st International Conference on Neural Information Processing Systems, Dec. 2017, pp.1782–1792. DOI: 10.5555/3294771.3294941.
    [48]
    Zhang H Y, Cai J J, Xu J J, Wang J. Pretraining-based natural language generation for text summarization. In Proc. the 23rd Conference on Computational Natural Language Learning, Nov. 2019, pp.789–797. DOI: 10.18653/v1/K19-1074.
    [49]
    Yan R. i, Poet: Automatic poetry composition through recurrent neural networks with iterative polishing schema. In Proc. the 25th International Joint Conference on Artificial Intelligence, July 2016, pp.2238–2244. DOI: 10.5555/3060832.3060934.
    [50]
    Shen L, Guo X Y, Chen M. Compose like humans: Jointly improving the coherence and novelty for modern Chinese poetry generation. In Proc. the 2020 International Joint Conference on Neural Networks, July 2020. DOI: 10.1109/IJCNN48605.2020.9206888.
    [51]
    Deng L M, Wang J, Liang H M, Chen H, Xie Z Q, Zhuang B J, Wang S J, Xiao J. An iterative polishing framework based on quality aware masked language model for Chinese poetry generation. In Proc. the 34th AAAI Conference on Artificial Intelligence, Feb. 2020, pp.7643–7650. DOI: 10.1609/aaai.v34i05.6265.
    [52]
    Shao Y Z, Shao T, Wang M H, Wang P, Gao J. A sentiment and style controllable approach for Chinese poetry generation. In Proc. the 30th ACM International Conference on Information & Knowledge Management, Oct. 2021, pp.4784–4788. DOI: 10.1145/3459637.3481964.
    [53]
    Ormazabal A, Artetxe M, Agirrezabal M, Soroa A, Agirre E. PoeLM: A meter- and rhyme-controllable language model for unsupervised poetry generation. In Proc. the Findings of the Association for Computational Linguistics, Dec. 2022, pp.3655–3670. DOI: 10.18653/v1/2022.findings- emnlp.268.
    [54]
    Cho K, van Merriënboer B, Gulcehre C, Bahdanau D, Bougares F, Schwenk H, Bengio Y. Learning phrase representations using RNN encoder-decoder for statistical machine translation. In Proc. the 2014 Conference on Empirical Methods in Natural Language Processing, Oct. 2014, pp.1724–1734. DOI: 10.3115/v1/D14-1179.
    [55]
    Mikolov T, Chen K, Corrado G, Dean J. Efficient estimation of word representations in vector space. arXiv: 1301.3781, 2013. https://arxiv.org/abs/1301.3781, Sept. 2023.
    [56]
    Zhao H K, Cheng Y H, Zhang X, Zhu H S, Liu Q, Xiong H, Zhang W. What is market talking about? Market-oriented prospect analysis for entrepreneur fundraising. IEEE Trans. Knowledge and Data Engineering, 2023, 35(6): 6489–6503. DOI: 10.1109/TKDE.2022.3174336.
    [57]
    Papineni K, Roukos S, Ward T, Zhu W J. BLEU: A method for automatic evaluation of machine translation. In Proc. the 40th Annual Meeting on Association for Computational Linguistics, July 2002, pp.311–318. DOI: 10.3115/1073083.1073135.
  • Related Articles

    [1]Yu Zhang, Tong Yu. Mining Trust Relationships from Online Social Networks[J]. Journal of Computer Science and Technology, 2012, 27(3): 492-505. DOI: 10.1007/s11390-012-1238-8
    [2]Xi-Shun Zhao, Yu-Ping Shen. Comparison of Semantics of Disjunctive Logic Programs Based on Model-Equivalent Reduction[J]. Journal of Computer Science and Technology, 2007, 22(4): 562-568.
    [3]Xiu-Li Sun, Wen-Yin Zhang, Jin-Zhao Wu. Event-Based Operational Semantics and a Consistency Result for Real-Time Concurrent Processes with Action Refinement[J]. Journal of Computer Science and Technology, 2004, 19(6).
    [4]LI Layuan, LI Chunlin. A Semantics-Based Approach for Achieving Self Fault-Tolerance of Protocols[J]. Journal of Computer Science and Technology, 2000, 15(2): 176-183.
    [5]Liu Weiyi, Yao Hong. A Logical Design Method for Relational Databases Based on Generalization and Aggregation Semantics[J]. Journal of Computer Science and Technology, 1997, 12(3): 252-262.
    [6]Cao Cungen, Liu Wei. A Three-Stage Knowledge Acquisition Method[J]. Journal of Computer Science and Technology, 1995, 10(3): 274-280.
    [7]Hock C. Chan. Translational Semantics for a Conceptual Level Query Language[J]. Journal of Computer Science and Technology, 1995, 10(2): 175-187.
    [8]Zhu Mingyuan. Two Congruent Semantics for Prolog with CUT[J]. Journal of Computer Science and Technology, 1990, 5(1): 82-91.
    [9]Wang Wenping, Wang Jiaye. A New Method for Generating Conic Sections[J]. Journal of Computer Science and Technology, 1989, 4(2): 155-162.
    [10]Tang Tonggao, Zhao Zhaokeng. Stack Method in Program Semantics[J]. Journal of Computer Science and Technology, 1987, 2(1): 51-63.
  • Others

Catalog

    Article views (281) PDF downloads (26) Cited by()
    Related

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return