Watson Extraction Extensions
The Watson Extraction Extensions uses the Watson Document Understanding technology to define the classes that extract structured document content into plain text, markdown, and JSON annotations.
- Display name
- Watson Extraction Extensions
- Installation type
- Optional
- Predecessors
- None
- Required by
- None
Property templates
The Watson Extraction Extensions add the following property templates.
| Property template | Type | Description |
|---|---|---|
| TxeWatsonServiceUrl | String | The service URL at which the handler can connect to either a Watson Document Understanding instance, or a watsonx.ai Text Extraction SaaS service. |
| TxeWatsonxApiKey | Binary | The API key for a watsonx.ai Text Extraction SaaS service. The system stores the key encrypted as UTF-8 encoded bytes. |
| TxeWatsonxSpaceGuid | String | The Space Id of the watsonx.ai deployment space hosting the watsonx.ai Text Extraction service. |
| TxeICOSUrl | String | The URL for the IBM Cloud Object Storage bucket that the watsonx.ai Text Extraction SaaS service uses. |
| TxeICOSApiKey | Binary | The API key for the IBM Cloud Object Storage bucket that the watsonx.ai Text Extraction SaaS service uses. The system stores the key encrypted as UTF-8 encoded bytes. |
| TxeICOSConnectionGuid | String | The Connection Id for the IBM Cloud Object Storage bucket that the watsonx.ai Text Extraction SaaS service uses. |
| TxeWatsonProcessingMode | String | Either "standard" or "high_quality". The addon initializes the value to "high_quality". |
| TxeWatsonOcrMode | String | One of "enabled", "disabled" or "forced". By default, the property has no value, implying disabled. |
| TxeWatsonOcrLanguages | String | An optional list of language identifiers to use when trying to OCR content. The addon initializes the list with a single entry "en_hw". |
| TxeWatsonDesiredMetadataTypes | String | Any combination of "kvps", "tables" and "markdown", specifying the desired metadata types that the system produces. An empty list implies that the system produces only the text extract. |
Objects
The Watson Extraction Extensions create the following objects:
| Object | Type | Description |
|---|---|---|
| TxeWatsonxExtractionAction | CmContentConversionAction | This action item is the object that invokes the Watson extraction handler for enhanced extraction. |
Classes
The Watson Extraction Extensions add the following classes.
| Class | Superclass | Description |
|---|---|---|
| TxeWatsonxExtractionAction | ReplicableClassDefinition | This class defines the Watson extraction action for enhanced extraction. |