At the intersection of creative text generation and literary theory, this study explores the role of literary metaphor and its capacity to generate a range of meanings. In this regard, literary metaphor is vital to the development of any particular language. To investigate whether the inclusion of original figurative language improves textual quality, we trained an LSTM-based language model in Afrikaans. The network produces phrases containing compellingly novel figures of speech. Specifically, the emphasis falls on how AI might be utilised as a defamiliarisation technique, which disrupts expected uses of language to augment poetic expression. Providing a literary perspective on text generation, the paper raises thought-provoking questions on aesthetic value, interpretation and evaluation.
Evaluating the Creativity of LLMs in Persian Literary Text Generation
Tehran Institute for Advanced Studies, Khatam University, Iran(德黑兰高级研究学院,卡坦大学,伊朗);School of Electrical and Computer Engineering, College of Engineering, University of Tehran, Tehran, Iran(电气与计算机工程学院,工程学院,德黑兰大学,德黑兰,伊朗)
CMDAG: A Chinese Metaphor Dataset with Annotated Grounds as CoT for Boosting Metaphor Generation
University of California, San Diego(加利福尼亚大学圣迭戈分校);Massachusetts Institute of Technology(麻省理工学院);HKUST(香港科技大学);University of Manchester(曼彻斯特大学);Institute of Computing Technology, Chinese Academy of Sciences(中国科学院计算技术研究所);University of Waterloo(滑铁卢大学);Multimodal Art Projection Research Community(多模态艺术投影研究社区)
Sapienza University of Rome Department of Computer, Control and Management Engineering(罗马大学Sapienza计算机、控制与管理工程系);DEXAI – Icaro Lab(DEXAI – Icaro实验室);University of Rome Tor Vergata(罗马大学Tor Vergata);Sant’Anna School of Advanced Studies(Sant’Anna高级研究学校)
Beyond Reproduction: A Paired-Task Framework for Assessing LLM Comprehension and Creativity in Literary Translation
Natural Language Learning Group (NLLG) School of Business Informatics and Mathematics University of Mannheim(曼海姆大学自然语言学习组(NLLG)商学院信息与数学学院);University of Technology Nuremberg (UTN), Department Engineering(纽伦堡技术大学(UTN)工程系);University of Gent, Department of Translation, Interpreting and Communication(根特大学翻译、口译与传播系);University of Aberdeen, Department of Computing Science(阿伯丁大学计算科学系);Natural Language Learning and Generation (NLLG) Lab(自然语言学习与生成(NLLG)实验室)
Generating similes effortlessly like a Pro: A Style Transfer Approach for Simile Generation
Information Sciences Institute, University of Southern California(南加州大学信息科学研究所);Columbia University(哥伦比亚大学);University of California, Los Angeles(加利福尼亚大学洛杉矶分校)
文学比喻,从诗歌到故事,是人类想象和交流的关键。比喻性的语言,如一个比喻,超越了简单的表达,给读者新的见解和灵感。在本文中,我们解决的问题,产生的。生成一个概念需要正确理解两个概念之间的属性的有效映射。为此,我们首先提出了一种方法来自动构建一个平行的语料库,通过转换大量的明喻从Reddit收集到他们的字面对应使用结构化的常识知识。然后,我们建议微调一个预先训练的序列到序列模型,BART~ \cite{lewis2019bart},对字面量-语义对获得泛化能力,这样我们就可以生成新的明喻给定一个字面量的句子。实验表明,我们的方法产生$88\%$新的明喻,不共享属性的训练数据。一独立集的文字语句的人类评价表明,我们的模型生成的明喻优于两个文学专家\textit{37\%} \footnote{We average 32.6\% and 41.3\% for 2 humans.}的时代,和三个基线系统,包括最近的隐喻生成模型\textit{71\%} \footnote{We average 82\% ,63\% and 68\% for three baselines.}的时代,当成对比较。\footnote{The simile in the title is generated by our best model. Input: Generating similes effortl…
Exploring Perspectives on the Impact of Artificial Intelligence on the Creativity of Knowledge Work: Beyond Mechanised Plagiarism and Stochastic Parrots
Microsoft Research(微软研究院);University of Cambridge(剑桥大学);University College London(伦敦大学学院)