How to use the Translation Memory Cleanup Tool.
Translation memory cleanup should retain context and approval history. Identical source text can legitimately have different translations in different domains, so ambiguous duplicates need review rather than blind merging.
Make the workflow fit your task.
Compare source, target, language, domain and interface context before identifying duplicates. Preserve approval status and provenance, separating conflicting translations from legitimate contextual variants. Propose merges only when their meaning and metadata are compatible.
- What you provide
- Small bilingual segment table and cleanup rules.
- What you get
- Duplicate, conflicting and malformed segment report.
See the input and the result.
Illustrative input and output · a teaching example, not a live WebAct run
Example input
Two entries translate Open into French as Ouvrir for a button and Ouvert for a status label.
Completed example
Keep both translations with their interface contexts. They are not interchangeable duplicates: one is an action, the other a state. Standardize metadata without erasing the distinction.
Load this input into the prompt, then copy it to WebAct to try the task. Your result may differ from the illustration.
Decisions and troubleshooting.
Should identical source words always share one translation-memory entry?
No. A word such as Open can describe an action or a state, requiring different translations and context.
Why did cleanup make a previously correct translation unsuitable for a button?
Context-specific entries may have been merged. Restore the action or status distinction and retain it in metadata.
Try it with your own source.
Replace the example with your material in the task prompt. Keep the requirements you need, then copy the task into WebAct.
Customize and copy the task ↑