{"slug": "ai-assisted-cultural-heritage-dissemination-comparing-nmt-and-glossary-augmented", "title": "AI-assisted cultural heritage dissemination: Comparing NMT and glossary-augmented LLM translation in rock art documents", "summary": "A study presented at the 26th Annual Conference of the European Association for Machine Translation (EAMT) found that glossary-augmented prompting of the Gemini LLM achieved the highest exact-match terminology accuracy (81.4%) in translating a Spanish rock art text into English, compared with 69.1% for Gemini with a basic prompt and 64.4% for DeepL, while preserving overall quality (mean DA 85.3 vs. 85.2 for Gemini-Simple and 80.3 for DeepL). The authors, Vicent Briva-Iglesias and María Ferre Fernández, conclude that glossary-augmented prompting is a low-overhead way to improve terminology control in cultural-heritage translation.", "body_md": "##### Abstract\n\nCultural 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: (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 lightweight retrieval of term pairs). Using PEARMUT, we conduct a human evaluation via (i) multi-way Direct Assessment (0–100) on 91 segments (1,743 Spanish words) 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) and outperforming DeepL (80.3). These results show that glossary-augmented prompting is a low-overhead way to improve terminology control in cultural-heritage translation, provided that institutions maintain minimal terminology resources and lightweight evaluation procedures.- Anthology ID:\n- 2026.eamt-1.49\n- Volume:\n[Proceedings of the 26th Annual Conference of the European Association for Machine Translation (Volume 1)](/volumes/2026.eamt-1/)- Month:\n- June\n- Year:\n- 2026\n- Address:\n- Tilburg, The Netherlands\n- Editors:\n[Dimitar Shterionov](/people/dimitar-shterionov/unverified/),[Eva Vanmassenhove](/people/eva-vanmassenhove/),[Mirella De Sisto](/people/mirella-de-sisto/unverified/),[Fred Blain](/people/frederic-blain/),[Javad Pourmostafa Roshan Sharami](/people/javad-pourmostafa-roshan-sharami/),[Lisa Lepp](/people/lisa-lepp/unverified/),[Chiara Manna](/people/chiara-manna/unverified/),[Argentina Anna Rescigno](/people/argentina-anna-rescigno/),[Alina Karakanta](/people/alina-karakanta/unverified/),[Ayla Rigouts Terryn](/people/ayla-rigouts-terryn/),[Manuel Lardelli](/people/manuel-lardelli/),[Natalia Resende](/people/natalia-resende/unverified/),[Elena Murgolo](/people/elena-murgolo/unverified/),[Janiça Hackenbuchner](/people/janica-hackenbuchner/),[Anna Zaretskaya](/people/anna-zaretskaya/unverified/),[Miquel Esplà-Gomis](/people/miquel-espla-gomis/unverified/),[Thierry Etchegoyhen](/people/thierry-etchegoyhen/unverified/),[Dagmar Gromann](/people/dagmar-gromann/),[Rachel Bawden](/people/rachel-bawden/),[Barry Haddow](/people/barry-haddow/unverified/),[Sara Szoc](/people/sara-szoc/unverified/),[Mikel Forcada](/people/mikel-l-forcada/),[Helena Moniz](/people/helena-moniz/)- Venue:\n[EAMT](/venues/eamt/)- SIG:\n- Publisher:\n- European Association for Machine Translation\n- Note:\n- Pages:\n- 774–786\n- Language:\n- URL:\n[https://aclanthology.org/2026.eamt-1.49/](https://aclanthology.org/2026.eamt-1.49/)- DOI:\n- Cite (ACL):\n- Vicent Briva-Iglesias and María Ferre Fernández. 2026.\n[AI-assisted cultural heritage dissemination: Comparing NMT and glossary-augmented LLM translation in rock art documents](https://aclanthology.org/2026.eamt-1.49/). In*Proceedings of the 26th Annual Conference of the European Association for Machine Translation (Volume 1)*, pages 774–786, Tilburg, The Netherlands. European Association for Machine Translation. - Cite (Informal):\n[AI-assisted cultural heritage dissemination: Comparing NMT and glossary-augmented LLM translation in rock art documents](https://aclanthology.org/2026.eamt-1.49/)(Briva-Iglesias & Fernández, EAMT 2026)- PDF:\n[https://aclanthology.org/2026.eamt-1.49.pdf](https://aclanthology.org/2026.eamt-1.49.pdf)", "url": "https://wpnews.pro/news/ai-assisted-cultural-heritage-dissemination-comparing-nmt-and-glossary-augmented", "canonical_source": "https://aclanthology.org/2026.eamt-1.49/", "published_at": "2026-08-24 00:00:00+00:00", "updated_at": "2026-08-27 08:19:33.224350+00:00", "lang": "en", "topics": ["large-language-models", "natural-language-processing", "ai-research"], "entities": ["European Association for Machine Translation", "Gemini", "DeepL", "PEARMUT", "Vicent Briva-Iglesias", "María Ferre Fernández"], "alternates": {"html": "https://wpnews.pro/news/ai-assisted-cultural-heritage-dissemination-comparing-nmt-and-glossary-augmented", "markdown": "https://wpnews.pro/news/ai-assisted-cultural-heritage-dissemination-comparing-nmt-and-glossary-augmented.md", "text": "https://wpnews.pro/news/ai-assisted-cultural-heritage-dissemination-comparing-nmt-and-glossary-augmented.txt", "jsonld": "https://wpnews.pro/news/ai-assisted-cultural-heritage-dissemination-comparing-nmt-and-glossary-augmented.jsonld"}}