Segmentation rules

Segmentation rules define the logic used by translation tools – such as Trados – to identify sentence boundaries based on punctuation, capitalization, formatting or language-specific patterns. For example, a period followed by a space and a capital letter typically marks the end of a segment in English, while abbreviations like “e.g.” or “Dr.” are exceptions that should not trigger segmentation.

Accurate segmentation rules ensure that translation memory (TM) systems can match and reuse content effectively. Poorly configured rules can break sentences incorrectly, lose context or reduce the accuracy of fuzzy matching and ICE matches. At RWS, segmentation rules are optimized for each project and language to preserve grammatical and contextual integrity. By combining automation in Trados with human linguistic expertise, RWS ensures that every translation memory is built on correctly segmented, high-quality data – improving consistency, speed and accuracy across projects.