Last Updated on

What’s New in Transact 4.1


Accuracy | Multidimensional classification

Ephesoft Transact 4.1 introduced the new patented multidimensional classification mechanism to improve accuracy of classification. This mechanism classifies documents across various dimensions and then combine the score of each dimension to give better accuracy and confidence. This classification mechanism works on supervised learning. It uses the learning to classify the pages in the given batch.

Ephesoft Transact 4.5 enhanced this classification mechanism and includes updates for this document.


Introduction to Multidimensional Classification

Overview of Multidimensional Classification

Effective with Ephesoft Transact Version 4.1.1.0, and continuing with later releases, Multidimensional Classification is available as a plugin mechanism within the Page Process module. This patented mechanism classifies documents across various dimensions, then combines the score of each dimension, resulting in better accuracy and confidence in document learning.

Comparing Multidimensional and Search Classification

Multidimensional Classification is a better approach to classifying documents, when compared to the older Search Classification, for the following reasons:

  • Search Classification was the original method of page classification with Ephesoft Transact. This older method continues to work well, but it is a less accurate approach. The classification accuracy is comparatively lower, when compared to the newer Multidimensional Classification.
  • Multidimensional Classification (this feature) contains newer algorithms that perform classification more accurately. This feature also improves document assembly.

Note: Transact 4.5.0.0, released in March 2018, further enhanced this feature by improving the proprietary classification algorithms and improving the user interface in the Page Process module. Additional information about these enhancements is available in the topic Enhancements in Ephesoft Transact Release 4.5.

Configuring Multidimensional Classification

This section describes how to configure, enable or disable Multidimensional Classification within a batch class.

Configuring Multidimensional Classification entails the following specific tasks:

  • Adding and enabling the MULTIDIMENSIONAL_CLASSIFICATION plugin in the Page Process module
    • This plugin is included by default in the BC1 batch class.
    • This plugin must be added manually to a batch class that is not based on the BC1 template batch class.
    • The default mode of this plugin is OFF.
  • Disabling the SEARCH_CLASSIFICATION plugin in the Page Process module, as needed
  • Configuring the DOCUMENT_ASSEMBLER plugin in the Document Assembly module to work with multidimensional classification

The following snapshot illustrates the plugin configurations for Multidimensional Classification. The administrator makes these configurations from the Batch Class Management screen.

Plugin requirements for Multidimensional Classification, Ephesoft Transact Release 4.5

 

Using Multidimensional Classification — Once configured, the Transact administrator trains at least one document for each document type with Multidimensional Classification, using the Learn button of the Document Types screen.

Configuring the Multidimensional Classification Plugin in the Page Process Module

Purpose — This topic describes how to add and enable the MULTIDIMENSIONAL_CLASSIFICATION_PLUGIN in the Page Process workflow module. This procedure also disables or removes the SEARCH_CLASSIFICATION feature.

Note: If you plan to use Multidimensional Classification to create classification results from document learning, and to perform document assembly, Ephesoft recommends disabling or removing the older Search Classification plugin.

Prerequisites — Ephesoft Transact must be operationally deployed and the user must have Administrator rights.

Procedure — Perform these steps to add, enable, disable or remove the MULTIDIMENSIONAL_CLASSIFICATION_PLUGIN within the Page Process module. This procedure also removes or disables the SEARCH_CLASSIFICATION plugin.

1. Launch Ephesoft Transact and navigate to Administrator > Batch Class Management. Provide login parameters as prompted.

2. Either select an existing batch class and click Open, or create a new batch class.
To create a batch class:

  • You can create a new batch class from scratch.
  • You can copy or import an existing batch class, then modify it.

Refer to one or more of the following articles for additional information about batch class creation, setup and configuration:

  • How to Create/Copy a New Batch Class

https://wiki.ephesoft.com/how-to-createcopy-a-new-batch-class-2

  • Batch Class Import/Export

https://wiki.ephesoft.com/batch-class-importexport

  • Productivity | Global Batch Class Management

https://wiki.ephesoft.com/home/documentation/global-batch-class-management

3. To add the MULTIDIMENSIONAL_CLASSIFICATION_PLUGIN to the batch class, perform these steps:

a. Within the batch class, navigate to Modules > Page Process. The Plugin Configuration screen for this module displays Associated Plugins and Selected Plugins.

b. Determine whether this plugin has been added to the module. The following snapshot illustrates this module without the plugin configured.

Sample Page Process Module without Multidimensional Classification Plugin

 

i. To add the plugin to the Page Process module, in the Plugin Configuration screen, select (highlight) the MULTIDIMENSIONAL_CLASSIFICATION_PLUGIN, and click the right-pointing arrow. This moves the plugin to the Selected Plugins field.

Associated Plugins for the Page Process Module, Release 4.5

 

ii. Adjust the position of this plugin in relation to the other plugins of this module. The following snapshot illustrates plugins that are common to the Page Process module.

Selected Plugins — Adjust the MULTIDIMENSIONAL_CLASSIFICATION_PLUGIN position as needful One typical sequence is shown above

 

iii. Click Apply to save the change, click Deploy to activate the change, and click Close to complete the configuration and exit the screen. The following message appears notifying that the plugin has been added to the batch class.

Confirmation Prompt

 

This plugin now appears within the Page Process module and is ready to be activated.

Page Process Module with MULTIDIMENSIONAL_CLASSIFICATION_PLUGIN

 

4. To enable the MULTIDIMENSIONAL_CLASSIFICATION_PLUGIN for the batch class, select (highlight) this plugin within the Page Process module. The Plugin Configuration screen appears on the right.

Dropdown Switch for Multidimensional Classification

 

a. On the right side of the screen, perform the following tasks:

— Select ON to enable this plugin. When this plugin is enabled, Transact will use this plugin to classify documents.

— Select OFF to disable this plugin. When this plugin is disabled, Transact will not use this plugin to classify documents.

b. Click Apply to save the changes. Click Deploy to enable the changes.

c. Click Close to exit the Plugin Configuration screen.

 5. Remove or disable the SEARCH_CLASSIFICATION plugin in the Page Process module, for the batch class that uses MULTIDIMENSIONAL_CLASSIFICATION_PLUGIN.

a. Navigate to the Page Process module. The Associated Plugins and Selected Plugins for the Page Process module appear on the right. This plugin will appear in the Page Process module, as illustrated with the following snapshot:

SEARCH_CLASSIFICATION Plugin in the Page Process Module

 

b. To disable the SEARCH_CLASSIFICATION plugin, select (highlight) this plugin. The Plugin Configuration screen appears. Select OFF from the dropdown menu.

Search Classification Switch

 

c. To remove the SEARCH_CLASSIFICATION plugin, select this plugin from the Selected Plugins field on the right. Use the left-facing arrow button to move this plugin to the Associated Plugins field on the left.

d. Click Apply to save this change. Click Deploy to activate this change.

6. Retrain Ephesoft Transact with Learn Files for the document type.

When you disable the Search Classification plugin, you must retrain the batch class with one or more documents using Multidimensional Classification. This is required because Multidimensional Classification generates its own model for classifying the data.

— Use the Learn Files button in the Batch Class Management screen.

— Refer to the separate topic, Document Learning with Multidimensional Classification.

Setting the Classification Type for the DOCUMENT_ASSEMBLER Plugin

Purpose — This topic describes how to select Multidimensional Classification as the classification type. Make this configuration in the DOCUMENT_ASSEMBLER plugin, which is located in Document Assembly module.

Prerequisites — Ephesoft Transact must be operationally deployed and the user must have Administrator rights. Also, the MULTIDIMENSIONAL_CLASSIFICATION_PLUGIN must be configured and enabled.

Procedure — Perform the following steps to configure the DOCUMENT_ASSEMBLER plug for use with the Multidimensional Classification plugin:

  1. Launch Ephesoft Transact and navigate to Administrator > Batch Class Management. Provide login parameters as prompted.
  2. Open the batch class in which the Multidimensional Classification plugin is enabled. Select the batch class, then click Open.
  3. Navigate to the Document Assembly module and select the DOCUMENT_ASSEMBLER plugin.
Document Assembly Module

 

The Plugin Configuration screen for DOCUMENT_ASSEMBLER plugin appears.

DOCUMENT_ASSEMBLER Plugin Configuration, Release 4.5

 

4. Select MultidimensionalClassification from the DA Classification Type dropdown list.

5. Click Apply and Deploy from the toolbar on top of the screen. The system provides confirmation windows for both Apply and Deploy.

6. Click Close to return to the Batch Class Management screen.

Removing or Disabling the Search Classification Plugin

Purpose — This topic describes how to disable or remove the SEARCH_CLASSIFICATION feature, which you should do when you enable Multidimensional Classfication.

Note: If you plan to use Multidimensional Classification to create classification results from document learning, and to perform document assembly, Ephesoft recommends disabling or removing the older Search Classification plugin.

Prerequisites — Ephesoft Transact must be operationally deployed and the user must have Administrator rights.

Procedure — Perform these steps to add, enable, disable or remove the MULTIDIMENSIONAL_CLASSIFICATION_PLUGIN within the Page Process module. This procedure also removes or disables the SEARCH_CLASSIFICATION plugin.

1. Launch Ephesoft Transact and navigate to Administrator > Batch Class Management. Provide login parameters as prompted.

2. Either select an existing batch class, and click Open, or create a new batch class. To create a batch class:

3. Remove or disable the SEARCH_CLASSIFICATION plugin in the Page Process module, for the batch class that uses MULTIDIMENSIONAL_CLASSIFICATION_PLUGIN.

a. Navigate to the Page Process module. The Associated Plugins and Selected Plugins for the Page Process module appear on the right. This plugin will appear in the Page Process module, as illustrated with the following snapshot:

SEARCH_CLASSIFICATION Plugin in the Page Process Module

 

b. To remove the SEARCH_CLASSIFICATION plugin, select this plugin from the Selected Plugins field on the right. Use the left-facing arrow button to move this plugin to the Associated Plugins field on the left.

c. To disable the SEARCH_CLASSIFICATION plugin, select (highlight) this plugin. The Plugin Configuration screen appears. Select OFF from the dropdown menu.

Search Classification Switch

 

d. Click Apply to save this change. Click Deploy to activate this change.

4. Retrain Ephesoft Transact with Learn Files for the document type, as applicable.

When you disable the Search Classification plugin, you must retrain the batch class with the documents using Multidimensional Classification. This is required because Multidimensional Classification generates its own model for classifying the data.

— Use the Learn Files button in the Batch Class Management screen.

— Refer to the separate topic, Document Learning with Multidimensional Classification.

Document Learning with Multidimensional Classification

Training Documents with Multidimensional Classification

The standard method of training a batch class for a document type applies to both Search Classification and Multidimensional classification.

The Multidimensional Classification mechanism works on supervised learning.

  • The plugin for Search Classification or Multidimensional Classification uses learning to classify the pages in the given batch. Such learning includes search classification.
  • The MULTIDIMENSIONAL_CLASSIFICATION_PLUGIN takes a sample of search classification and updates it during the learning process.

Within an open batch class, use the Learn File(s) button in the Batch Class Management screen. When a document type is trained with at least one file, the plugin creates a new file with the following name in the Batch Class folder:

  • BC<ID>-dimensions

Example: BC8-dimensions

In addition, the user can click Learn File(s) to update the learning that occurs during classification.

The following two snapshots illustrate the Learn File(s) button from an open batch class from the Batch Class Management screen.

Learn Files in the Document Types screen for a sample batch class, Ephesoft Transact Release 4.1.1.0

 

Learn Files in the Document Types screen for a sample batch class, Ephesoft Transact Release 4.5

 

Multidimensional Classification and Machine Learning

The MULTIDIMENSIONAL_CLASSIFICATION_PLUGIN feature is supported by machine learning for document types. In this case, if classification is done incorrectly during the initial process job, the plugin can learn the change that is required to correct the classification, so that it performs classification correctly with ensuing batch processing jobs.

Testing Classification of a Document Type with Multidimensional Classification

Purpose — This topic describes how to learn a new document to test for classification of a new document type.

Prerequisites — With an operational installation of Ephesoft Transact, the two primary requirements for this procedure are as follows

  • Previous configurations for the MULTIDIMENSIONAL_CLASSFICATION_PLUGIN, SEARCH_CLASSIFICATION and DOCUMENT_ASSEMBLER plugins must be complete before performing these steps. Refer to earlier topics in this document for additional information about configuring these two plugins:

Configuring the MULTIDIMENSIONAL_CLASSIFICATION_PLUGIN in the Page Process Module

Setting the Classification Type for the DOCUMENT_ASSEMBLER Plugin

  • Also, to complete this procedure, you must have sample documents available in electronic format, PDF or TIFF, for each document type that is to be tested.

Procedure — Perform the following steps to test classification of a document type, for the batch class in which multidimensional classification is enabled.

1. Launch Ephesoft Transact and navigate to Administrator > Batch Class Management. Provide login parameters as prompted.

2. Open the batch class in which the Multidimensional Classification plugin is enabled. Select the batch class, then click Open. The batch class opens with a list of document types.

3. To learn samples for a particular document type, upload sample documents.

  • For each document type, click Learn Files.
  • Select the document type. Then, click the Upload Test Classification File(s) link to select and upload a test image file for the document type.

The following message appears notifying you when the test file upload is complete. You can also drag and drop the sample image file to the Drag and Drop Files Here area below the Upload Test Classification File(s) link.

The system provides the following confirmation message for a sample upload.

Success confirmation dialog

 

Note: You can also drag and drop the sample image file to the Drag and Drop Files Here area below the Upload Test Classification File(s) link.

4. Navigate to the document type screen where you uploaded the test image file (in the previous step) and click Test Classification from the toolbar on top of the screen. The Test Classification screen displays.

Test Classification screen

 

5. Select an option from the Workflow drop-down list as described in the following summary:

 

 

 

 

 

 

 

 

 

ON — If the Workflow field is set to ON, then the Classification Types dropdown list is disabled. Test classification results are based on configurations within the batch class.

Test Classification dialog

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

OFF — If the Workflow is set to OFF, then test classification results are based on the selection you make from the Classification Types drop-down list.

Classification Types dropdown menu options

 

The various classification types available are as follows:

  • Search Classification
  • Barcode Classification
  • Image Classification
  • Automatic Classification
  • Keyword Classification
  • Multidimensional Classification

6. Click Classify. The Test Classification screen is updated with classification results as shown in the following image.

Test Classification screen with sample results

 

Enhancements in Ephesoft Transact Release 4.5

Learning with Multidimensional Classification in Release 4.5.0.0

Multidimensional classification can be described as “page classification learning combined with search classification learning.” This classification mechanism works on supervised learning.

• The user begins this process by importing documents, then moving to the Page Processing module and clicking Learn Files to update learning for page and search classification.

• The plugin uses learning to classify the pages in the given batch, and simultaneously learns search classification.

• The plugin takes the sample of search classification and makes corresponding updates to the page-classification learning.

Phases of Multidimensional Classification in Release 4.5.0.0

Click Learn Files to begin the learning phase. A single model file is kept per batch class. In the learning phase for a document, the system performs calculations and saves the model file. Each document page type (first page, middle page and last page) saves multiple types of information.

During the batch execution phase, the pages in the batch are classified according to learning.

When Transact performs auto-learning of the document type, along with the indexes for search classification, Transact also updates the model file for multidimensional classification.

During the document assembly phase, this plugin works in the same way as for search classification.