AI-assisted cultural heritage dissemination: Comparing NMT and glossary-augmented LLM translation in rock art documents

2026-05-14

AI Summary

This study addresses the challenge of inaccurate terminology translation in multilingual dissemination of specialized cultural heritage documentation—such as rock art literature—where limited resources often mislead non-specialists and hinder content reuse. To mitigate this, the authors propose a lightweight terminology-augmented prompting strategy that integrates domain-specific terms into the Gemini large language model via Retrieval-Augmented Generation (RAG), enhancing terminological consistency and translation quality without modifying the model architecture. The approach is evaluated against DeepL and baseline Gemini using both PEARMUT human assessment and constrained MQM terminology auditing. Results demonstrate that Gemini-RAG achieves 81.4% exact term-match accuracy and an overall translation quality score of 85.3, significantly outperforming DeepL (64.4%/80.3) and baseline Gemini (69.1%/85.2).

Abstract

Cultural heritage institutions increasingly disseminate research and interpretive materials globally, but multilingual dissemination is constrained by limited budgets and staffing. In terminology-dense domains such as rock art, translation quality depends on accurate, consistent specialised terms, and small lexical errors can mislead non-specialists and reduce reuse. We compare three English MT setups for a Spanish academic rock art text, focusing on simple, operationally feasible interventions rather than complex model-side modifications: (1) DeepL as a strong NMT baseline, (2) Gemini-Simple (LLM with a basic prompt), and (3) Gemini-RAG (the same LLM with glossary-augmented prompting via term-pair retrieval). Using PEARMUT, we conduct a human evaluation via (i) multi-way Direct Assessment (0--100) and (ii) targeted terminology auditing with a restricted MQM taxonomy. Gemini-RAG yields the highest exact-match terminology accuracy (81.4\%), versus Gemini-Simple (69.1\%) and DeepL (64.4\%), while preserving overall quality (mean DA 85.3 Gemini-RAG vs. 85.2 Gemini-Simple), outperforming DeepL (80.3). These results show that glossary-augmented prompting is a low-overhead way to improve terminology control in cultural-heritage translation if institutions maintain minimal terminology resources and lightweight evaluation procedures.

Problem

• cultural heritage
• machine translation
• terminology accuracy
• rock art
• multilingual dissemination

Innovation

• glossary-augmented prompting
• terminology accuracy
• large language models
• machine translation
• cultural heritage dissemination