Extract structured data from text using AWS Bedrock
| Property | Details |
|---|---|
| Purpose | Extract structured information from unstructured text |
| Usage | Transform free-form text into structured data formats |
| Category | Intelligent Assist |
| Type | AI Object |
| Security | Safe data extraction with validation |
| Parameter | Type | Description |
|---|---|---|
| Extract From | textarea |
The text content to extract data from |
| Model | modelSettings |
AI model for text analysis and extraction |
| Data | object |
Input data context and additional variables |
| Parameter | Type | Default | Description |
|---|---|---|---|
| Outputs | dynamicOutput2 |
- | Custom output fields to extract |
| Model Config | object |
- | Additional model configuration options |
{
extractedData: {
[fieldName: string]: any // Dynamic fields based on configured outputs
}
}
Returns structured data extracted from the input text
The Extract node provides intelligent data extraction:
| Feature | Description |
|---|---|
| Text Input | Process various types of unstructured text |
| Dynamic Outputs | Configure custom extraction fields and data types |
| Model Selection | Choose appropriate models for different extraction tasks |
Choose the Extract node when you need to:
| Aspect | Requirement | Impact |
|---|---|---|
| Text Quality | Clear, structured text content | Affects extraction accuracy and completeness |
| Output Definition | Well-defined extraction fields | Determines what data is captured |
| Model Selection | Text-capable models | Required for natural language processing |
| Factor | Description | Optimization |
|---|---|---|
| Text Length | Amount of text to process | Balance detail with processing efficiency |
| Field Count | Number of fields to extract | More fields increase processing time |
| Processing Time | Model inference time | Choose appropriate models for speed vs accuracy |
| Security Feature | Protection | Benefit |
|---|---|---|
| Input Validation | Validate text inputs | Prevent malicious content processing |
| Output Sanitization | Clean extracted data | Ensure safe data consumption |
| Data Validation | Validate extracted fields | Prevent incorrect data extraction |
Purpose: Extract billing information from invoices
Text: Invoice document content
Outputs: Vendor, Amount, Date, Invoice Number, Line Items
Result: Structured invoice data for processing
Purpose: Extract contact details from business cards or emails
Text: Business card scan or email signature
Outputs: Name, Email, Phone, Company, Address
Result: Structured contact information
Purpose: Extract key metrics from business reports
Text: Quarterly report content
Outputs: Revenue, Growth, Market Share, KPIs
Result: Structured business metrics
Purpose: Extract key information from news articles
Text: News article content
Outputs: Headline, Date, Author, Summary, Keywords
Result: Structured article metadata