Which metric family compares a summary with a reference summary?
Why D is the answer
ROUGE
ROUGE measures overlap with reference summaries.
In detail
- ROUGE measures overlap with reference summaries.
- It measures overlap.
Use case
An illustrative scenario, not a real incident.
A news team compares generated summaries with reference summaries, using ROUGE.
See it in different ways
flowchart TD
A{"What are you evaluating?"} -->|"Translation"| B["BLEU"]
A -->|"Summaries"| C["ROUGE"]
A -->|"RAG answers"| D["Faithfulness and relevance"]
A -->|"Open-ended quality"| E["Human or LLM judged rubric"]mindmap
root((Evaluation))
Translation
BLEU
Summaries
ROUGE
RAG
Faithfulness
Errors
Slice by topicflowchart LR S0["Generated summary"] --> S1["Reference summary"] --> S2["ROUGE"]
Key point: ROUGE compares summaries.
Why the other options are wrong
- A. p99 latency
- That measures latency.
- B. GPU utilization
- That measures hardware.
- C. Requests per second
- That measures load.