Trainable locator method types
Unlike the standard locator methods, group locator methods are based on trained data. You can create this trained data by training your documents or by using predefined knowledge bases. To train your documents, you need sample documents that show the system where fields are positioned on the document. This information is then trained into the project and used at runtime. Each trainable locator method has its own property window so you can adjust specific parameters and settings.
The following trainable locator method types are available:
| Icon | Evaluator Method Type | Description |
|---|---|---|
|
|
This trainable group locator extracts fields that are related to amounts, including the total. For example, net amount, total amount, tax rate, tax amount are all part of this group. Many of these fields are optional and need not be present on the invoice. |
|
|
|
This trainable group locator extracts fields related to the type of information typically found in invoice headers, such as invoice number, invoice date, customer ID and VAT ID. |
|
|
|
This trainable group locator extracts fields related to the supplier or vendor, such as the order number and order date. |
|
|
|
This trainable locator extracts table data from a document when enabled.
Note To train this locator method, manually enable training.
|
|
|
|
This trainable locator extract sdata from unstructured documents. Note You cannot train this locator using
online learning. Although this locator can collect training documents, you must train it manually, which can
be time consuming.
|
|
|
|
This evaluator trains the best alternatives from other locators in the project. |
|
|
|
This trainable group locator contains trainable fields for all relevant field data on a document. |
All trainable locators have test modes that help you fine-tune and configure locator settings. Test results are highlighted in the Document Viewer and displayed in an extraction result list , sorted by confidence.
In addition to these trainable locator method types, there are locator methods and evaluator methods available.