Coldmist-Lu/ErrorAnalysis_Prompt

:gift:[ChatGPT4MTevaluation] ErrorAnalysis Prompt for MT Evaluation in ChatGPT

27
/ 100
Experimental

This helps translation evaluators accurately assess the quality of machine translations. By providing a source text and a machine-translated output, it identifies and categorizes translation errors, then assigns a comprehensive quality score. The tool is ideal for anyone working with machine translation, such as translation project managers or linguists, who needs to quickly and reliably gauge translation quality.

No commits in the last 6 months.

Use this if you need a reliable and explainable way to evaluate machine translation quality, emulating human-like assessment through detailed error analysis.

Not ideal if you lack access to a large language model (LLM) API, as it's a prompting-based technique requiring a chat interface to an LLM.

machine-translation linguistics translation-quality-assurance localization translation-project-management
No License Stale 6m No Package No Dependents
Maintenance 2 / 25
Adoption 9 / 25
Maturity 8 / 25
Community 8 / 25

How are scores calculated?

Stars

92

Forks

5

Language

Ruby

License

Last pushed

Oct 14, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/prompt-engineering/Coldmist-Lu/ErrorAnalysis_Prompt"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.