In recent years, with the rapid development of deep learning, natural language processing has achieved great progress in many aspects. In the field of text generation, classical Chinese poetry, as an ...important part of Chinese culture, also attached growing attention. However, the existing researches on neural-network-based classical Chinese poetry generation ignore the semantics contained in Chinese words. A sentence in Chinese is a sequence of characters without spaces, and thus it is of great significance to segment the sentence properly for understanding the original text correctly. Therefore, supposing that the model knows how to segment the sentence, the meaning of the sentence will be more accurately understood. In this paper, we propose a novel model, namely WE-Transformer (Word-Enhanced Transformer), to generate classical Chinese poetry from vernacular Chinese in a supervised approach, which incorporates external Chinese word segmentation knowledge. Our model learns word semantics based on character embeddings by bidirectional LSTM and enhances the quality of generated classical poems based on the Transformer with extra word encoders. Compared to the baselines and state-of-the-art models, our experiments on automatic and human evaluations have demonstrated that our method can bring better performance.
The computer generation of poetry has been studied for more than a decade. Generating poetry on a human level is still a great challenge for the computer-generation process. We present a novel ...Transformer-XL based on a classical Chinese poetry model that employs a multi-head self-attention mechanism to capture the deeper multiple relationships among Chinese characters. Furthermore, we utilized the segment-level recurrence mechanism to learn longer-term dependency and overcome the context fragmentation problem. To automatically assess the quality of the generated poems, we also built a novel automatic evaluation model that contains a BERT-based module for checking the fluency of sentences and a tone-checker module to evaluate the tone pattern of poems. The poems generated using our model obtained an average score of 9.7 for fluency and 10.0 for tone pattern. Moreover, we visualized the attention mechanism, and it showed that our model learned the tone-pattern rules. All experiment results demonstrate that our poetry generation model can generate high-quality poems.
Classical Chinese poetry and English poetry show huge differences, mainly with regard to literary imagery, sense, rhythm, sound, genre, and format. Hence, English translation of classical Chinese ...poetry based on equivalence is not an option. This article intends to explore translation as a creative, cross‐cultural practice. We shall suggest that “mirror‐shared reality” may play a critical role in this practice. This notion refers to an intermediate construction made by the translators with the aim of bridging different “realities” of classical Chinese poetry and English poetry with regard to knowledge, tradition, and reference. The construction of a “mirror‐shared reality” is a type of creative practice based on both realities, not just one of them. By selecting some of their relevant shared features, the two realities will mirror each other in certain essential aspects and produce an aesthetic experience among English readers theoretically akin to the experience among their Chinese counterparts. Our discussion is based on a number of English translations of classical Chinese poems, in particular translations by Xu Yuanchong (许渊冲 Xǔ Yuānchōng), building on his own theory, “Threefold‐Beauty” (三美论 sān měi lùn).
This article explores the possibility of introducing classical Chinese poetry to primary-aged children in China through educational drama. Today young people in China generally regard classical ...poetry as cultural relic which is remote and irrelevant to their lives. Referring to writings of Chia-ying Yeh and Stephen Owen, I argue that the essence of classical poetry study is to evoke the reader's literary imagination and emotional response to the poetic discourse. Drawing from reception theory, I suggest that drama as a pedagogy can connect with the aesthetic discourses and to help children engage with the text. This article presents theoretical explanations and practical descriptions on how the drama approaches are designed to provide children with the contextual references that allow them to relate to the poem at their personal levels of experience; and to help them feel, in an embodied form, the beauty of poetic language.
This paper aims to develop a feasible way to recognize the style of classical Chinese poetry with computers. To this end, the authors explored the connectionism in neuroscience, and explained the ...cognitive word embedding with the convolutional neural network (CNN). On the one hand, the genetic algorithm was adopted to extract keywords from traditional hand-labelled and selected information; on the other hand, a novel computer learning method was proposed based on text-to-image (T2I) CNN for big data. The proposed method was contrasted with the traditional genetic algorithm of naive Bayes and information gain. The experimental results show that our method achieved better classification accuracy with less human intervention than the traditional genetic algorithm. Hence, the CNN-based method is feasible on big data, both in theory and practice. This cross-disciplinary practice sheds light on stylistics, literature engineering, poetry cognition and neural network projects.