Exploring Chinese Humor Generation: A Study on Two-Part Allegorical Sayings

arXiv:2403.10781v1 Announce Type: new
Abstract: Humor, a culturally nuanced aspect of human language, poses challenges for computational understanding and generation, especially in Chinese humor, which remains relatively unexplored in the NLP community. This paper investigates the capability of state-of-the-art language models to comprehend and generate Chinese humor, specifically focusing on training them to create allegorical sayings. We employ two prominent training methods: fine-tuning a medium-sized language model and prompting a large one. Our novel fine-tuning approach incorporates fused Pinyin embeddings …

abstract art arxiv capability challenges chinese community computational cs.ai cs.cl generate human humor language language models nlp paper part state study type understanding