Translation Memory (TM)
A Translation Memory (TM) is a database that stores previously translated text segments (source-target text pairs) and suggests them for reuse in future translation projects. When a new segment matches a stored entry (100% match) or shows high similarity (fuzzy match), the existing translation is proposed. TMs are a core component of modern CAT tools and contribute significantly to terminological consistency, time savings, and cost reduction.
In the context of IFU translation, a Translation Memory enables consistent reuse of regulatory-approved formulations across product families and revision cycles, ensuring quality while reducing the effort required for updates.
Frequently Asked Questions
How does a Translation Memory improve the quality of IFU translations?+
A Translation Memory ensures that identical or similar text segments are translated consistently – across all language versions and product variants. This is particularly critical for regulatory-sensitive formulations such as warnings, contraindications, and intended purpose, where inconsistencies can lead to regulatory objections.