Quality Evaluation of Technical Text Translation Based on the HACT Model

DOI: https://doi.org/10.62517/jhet.202615326

Author(s)

Daohan Li

Affiliation(s)

School of Humanities and Education, Guangzhou Huanan Business College, Guangdong, China

Abstract

The Human-AI Collaborative Translation (HACT) model has three core modules, which are collaborative pre-editing, collaborative mid-translation, and collaborative post-editing. These modules enable dynamic, full-process collaboration between human translators and artificial intelligence systems. However, the actual effectiveness of the HACT model in the translation of technical texts has yet to be systematically tested. To address this gap, this study employs DeepSeek as an empirical tool, selects a corpus of technical texts, and introduces the Multidimensional Quality Metrics (MQM) framework for error classification and quality assessment. By comparing the quality differences between raw machine translation outputs and HACT-mediated collaborative translations across four dimensions-terminology errors, logical errors, accuracy, and fluency-this study examines the extent to which the HACT model improves translation quality.

Keywords

HACT Model; Multidimensional Quality Metrics (MQM); Technical Texts; Translation Quality Assessment; DeepSeek

References

[1]Liu, C., Yang, C., & Wang, T. (2023). An overview of scientific text translation research from the perspective of relevance theory. Language and Culture Research, 26(1), 123-125. [2]Wen, Q., & Liang, M. (2024). Human-computer interaction negotiation competence: ChatGPT and foreign language education. Foreign Language Teaching and Research, 56(2), 286-296. [3]Liu, S., & Zhang, Y. (2025). From MTPE to HACT: A study on innovation in translation process driven by large language models. Foreign Language Education Research, 13(1), 18-26. [4]Wang, Y., & Zhang, Z. (2024). Concerns and solutions of ChatGPT AI translation. Chinese Translators Journal, (2), 95-102. [5]Multidimensional Quality Metrics Community. (n.d.). Multidimensional Quality Metrics (MQM). Retrieved June 5, 2026, from https://themqm.org/ [6]Wang, J., & Zhang, Z. (2026). Construction of human-AI collaborative translation model from the perspective of human-machine symbiosis. Shanghai Journal of Translators, 1(1), 13-19. [7]Wang, J., & Zhang, Z. (2025). A large language model-driven human-machine collaborative translation teaching model. Foreign Language World, (5), 69-76.