[LIVE WEBINAR] Context is Everything: Maximizing Large Language Models in Machine Translation through Contextual Enrichment
Members: Free
Non-Members: $75
Please log in to your GALA account to register. Don’t have an account yet?
Create one for free.
Machine translation has been proved to reduce the costs and delivery times in various translation scenarios. Despite achieving impressive quality, machine translation still can not compete with human translation, especially when the availability of context is limited. The rise of Large Language Models makes it possible to leverage context information in linguistic tasks. With that in mind we developed the XTM AI Smart Context feature which utilizes translation memory matches and carefully-designed prompts in LLM-powered machine translation. Evaluation results show that this approach is very promising.
The session will compare the metrics scores achieved by different GPT models against state-of-art machine translation engine.
Takeaways for the audience
1. When prompted with context, LLMs offer unprecedented machine translation quality
2. LLMs can truly help in production scenarios.
3. Linguistic data management is key - a TMS with AI capabilities is the tool of the future.
Host organization: XTM International
Event Speakers
Rafał Jaworski
XTM International
Rafał Jaworski, PhD, is an academic lecturer and scientist specializing in natural language processing techniques. His alma mater is Adam University in Poznań, Poland where he works at the Department of Artificial Intelligence (AI). Rafał’s scientific work concentrates around developing robust AI algorithms for the needs of computer-assisted translation. These include automatic lookup of linguistic resources and computer-assisted post editing. Apart from the research and teaching, he works as a linguistic AI expert at XTM International leading a team of young and talented developers who put his visions and ideas into practice.