You are working on a project to implement a custom translation model for a specific domain using Azure AI Translator service. The domain has a large amount of technical jargon and specialized vocabulary. Which approach would you take to improve the accuracy of the custom translation model?