Discover insights and relationships in text
Amazon Comprehend is a natural-language processing (NLP) service that uses machine learning to uncover information in unstructured data. Instead of combing through documents, the process is simplified and unseen information is easier to understand.
When Amazon Comprehend analyzes a document, it identifies the following:
- Detects entities based on built-in models
- Detects key phrases
- Determines the dominant language
- Detects personally identifiable information (PII)
- Determines the sentiment of the text
- Analyzes the syntax in the document
In addition to providing offsets of detected PII entities, a PII analysis job can also provide redactions of detected PII entities.
Beneath the analysis, Amazon Comprehend displays a confidence assessment of the results. Under Application Integration, it also shows the Request and Response JSON code for the API being used for each type of analysis.
If you created an endpoint for a custom model, the model is available for real-time analysis. If you use the endpoint for a custom classifier model, Amazon Comprehend classifies the input text according to the model’s categories or labels. If you use the endpoint for a custom entity recognizer, Amazon Comprehend analyzes the input text to detect the model’s entities.