Now that the accounts payable batch class exists, we must tell Ephesoft about the types of documents that this batch class will process. Each batch class can process many different document types, but we will begin by defining just one document type: an invoice.

Select the newly created batch class and click on the Open button.

When editing batch classes, Ephesoft displays a navigation tree on the left that allows the user to select the aspect of the configuration to be modified. The Document Types item is the default, so it should already be selected, as shown in the following screenshot:

The batch class document type list

Click on the Add button located on the top of the screen. A new row will appear in the Document Types table into which you can enter information about the new document type, as shown in the following screenshot:

New document type configuration

The following information can be provided when creating a new document type:

Traditional image capture systems require that scan operators place separator sheets between each document in a batch. The separator sheets tell the system where each document begins and ends and also indicate the document type. We will configure Ephesoft to separate the example accounts payable batch into individual documents and classify these documents as being invoices without the use of separator sheets. In order to accomplish this, we must train Ephesoft to recognize invoices by supplying an example.

First, find or create a blank invoice. If a blank invoice is not available, you can use an actual invoice with line items and header information, but this won't be as effective. We will discuss training and sample documents further in the next chapter.

Select the checkbox next to the invoice document type. Then, drag and drop the sample invoice to the area below the Upload Learn File(s) link.

Ephesoft will display a message while it learns...

Ephesoft stores information about a document in document-level fields. Further, Ephesoft stores information common to all the documents in a batch in batch-level fields. Typically, field content is extracted from a document, but the fields can be populated in a number of other ways. The field content can be manually entered by an operator, for instance, or populated from a database, or calculated programmatically on the basis of the values of other fields.

For the invoice document type, we would like to extract the customer number, invoice number, and invoice date. In order to accomplish this, we must first create the fields in the document type that we created. From the Batch Class Management interface, select and edit the Accounts Payable batch class. From the Document Types tree node, open the Invoice document type. The Index Fields item should be selected by default. Click on the Add button at the top of the user interface, which will create a new row in the table of...

Extraction is the process of automatically populating fields with text from a document. In the following example, the customer number follows a label with the text Customer Number. We can configure Ephesoft to extract any numeric text following Customer Number into the Customer Number field. This process is known as key/value extraction. The key is the label, and the value is the text to be extracted.

The sample invoice form

To define a key/value extraction rule, navigate to the KV Extraction Rule area of the menu on the left side of the batch class administration screen. Click on the Add button to open the key/value rule screen for a new rule.

In this screen, indicate on an actual invoice where the key and the value are located. Do this by dragging the invoice into the KV Test area at the bottom of the screen. The first page of the invoice will fill the right side of the screen, with two colored rectangles. Drag the green rectangle so that it surrounds the label, as...