Optimising ChatGPT for creativity in literary translation: A case study from English into Dutch, Chinese, Catalan and Spanish
By: Shuxiang Du , Ana Guerberof Arenas , Antonio Toral and more
Potential Business Impact:
Makes computer translations more creative and interesting.
This study examines the variability of Chat-GPT machine translation (MT) outputs across six different configurations in four languages,with a focus on creativity in a literary text. We evaluate GPT translations in different text granularity levels, temperature settings and prompting strategies with a Creativity Score formula. We found that prompting ChatGPT with a minimal instruction yields the best creative translations, with "Translate the following text into [TG] creatively" at the temperature of 1.0 outperforming other configurations and DeepL in Spanish, Dutch, and Chinese. Nonetheless, ChatGPT consistently underperforms compared to human translation (HT).
Similar Papers
ChatGPT as a Translation Engine: A Case Study on Japanese-English
Computation and Language
Makes computers translate Japanese to English better.
Has the Creativity of Large-Language Models peaked? An analysis of inter- and intra-LLM variability
Computation and Language
Computers aren't getting more creative, even the best ones.
ChatGPT as Linguistic Equalizer? Quantifying LLM-Driven Lexical Shifts in Academic Writing
Computation and Language
Helps scientists write better, especially non-native English speakers.