Evaluator method types
Evaluator locator methods evaluate data and then return the most appropriate alternative to the field.
The following evaluator method types are available:
|
Icon |
Evaluator Method |
Description |
Extraction Type |
|---|---|---|---|
|
|
The Address Evaluator takes the results from several locators and compares those results to records in a database. If the locator values match the records in the database, the Address Evaluator returns matching records. |
Group |
|
|
|
Use the Advanced Evaluator to return the best extraction results from multiple input locators for a set of subfields commonly found on invoice and purchase order documents. For each subfield, define evaluation steps and confidence thresholds for the input locators to find the correct result in a voting process. This evaluator provides comparisons for up to three input locators for typical invoice header data, such as invoice number, order date, different amounts, or other custom subfields. |
Group |
|
|
|
The Database Evaluator identifies a database record that matches the input data for a document. Unlike the Database Locator, that attempts to match a database record with recognition data from the full page or a smaller region, the Database Evaluator uses data from the results of other locators. |
Group |
|
|
|
The OCR Voting Evaluator compares the results an Advanced Zone Locator and selects the best result to save to the field. |
Group |
|
|
|
The Relation Evaluator locates the best alternatives from one locator based on their position on the document in relation to the best alternative of another locator. |
Single |
|
|
|
The Standard Evaluator compares the results from several different locators, and selects a set of results based on preset criteria. |
Single Table |
|
|
|
If you want quick and good generic extraction results with a relatively small training set, the highly optimizable Trainable Evaluator, is recommended. This evaluator is used to compare alternatives from other locators to determine which of those alternatives match a specific set of criteria. This evaluator relies on alternatives from the input locator. It also learns from false alternatives to improve training. |
Group |
All evaluators have test modes that help you fine-tune and configure settings. Test results are highlighted in the Document Viewer and displayed in an extraction result list as well, sorted by their confidence.
In addition to these evaluator method types, there are locator methods and trainable locator methods.
Related topics: