easyMultiple Choice
AIF-C01 Practice Question: A developer is building an application that…
A developer is building an application that translates customer support tickets from English to Spanish using Amazon Bedrock. They need to evaluate the quality of translations. Which automated metric is most appropriate for comparing the model's translations to professional human translations?
⚠ Common exam trap
The trap is that BERTScore and ROUGE sound more 'modern' and semantically aware, so candidates pick them — but the exam expects you to know BLEU is the canonical metric for machine translation specifically, while ROUGE is for summarization and BERTScore is general-purpose.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
BLEU
BLEU (Bilingual Evaluation Understudy) is specifically designed to evaluate machine translation quality by comparing n-gram overlap between the model's output and one or more professional human reference translations. It is the standard automated metric for translation tasks on Amazon Bedrock and in NLP generally.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Accuracy
Why it's wrong here
Accuracy is a classification metric counting exact-match predictions, so it cannot score a translation where no single token sequence is uniquely correct. It suits labelled classification tasks such as sentiment or intent detection. Translation quality needs BLEU, which measures n-gram precision against professional human reference translations.
- ✓
BLEU
Why this is correct
BLEU scores machine translation by measuring n-gram overlap between model output and reference human translations, making it the standard automated metric for this comparison. It requires no model-based judging and directly quantifies translation quality against professional references.
- ✗
BERTScore
Why it's wrong here
BERTScore compares contextual embeddings token by token, capturing semantic similarity but not the fluency or adequacy judgements that trained metrics like COMET or BLEURT provide against human references. It is tempting because it is automated and reference-based, and suits tasks where surface overlap matters less than meaning.
- ✗
ROUGE
Why it's wrong here
ROUGE measures n-gram overlap with reference summaries, so it cannot judge translation adequacy or word order across languages. It is designed for summarisation evaluation, where it would be the right metric. Translation comparison requires BLEU, METEOR or COMET, which score bilingual correspondence against human references.
Go deeper
Related to this question
About these practice questions
One of 862 original AIF-C01 practice questions on Courseiva, each with a full explanation and wrong-answer analysis — not exam dumps or protected exam content. Learn why practice questions differ from exam dumps →
JA
Written and reviewed by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
Last reviewed September 2026 · checked against the official Amazon Web Services exam blueprint
This AIF-C01 practice question is part of Courseiva's free Amazon Web Services certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the AIF-C01 exam.