Basic Information
| Field | Details |
|---|---|
| English name | Amazon Comprehend |
| Full name | Amazon Comprehend |
| Chinese description | 自然语言处理 |
| Japanese description | 自然言語処理 |
| Exam frequency | ⭐⭐ |
| Often confused with | Translate / Bedrock |
In one sentence
Use natural-language processing to identify sentiment, entities, key phrases, language, and topics in text.
Key points
- It fits analysis of feedback, tickets, reviews, and documents.
- The input must already be text; transcribe audio first with Transcribe.
- Comprehend understands text; it does not generate speech or translate languages.
Exam focus
- Sentiment, entities, key phrases, and text insights point to Comprehend.
Common pitfalls
- Do not choose by product name alone; confirm data type, latency, control, operations, and cost constraints.
Remember
Comprehend understands text; Transcribe turns speech into text.