Create Schema for Purchase Order Documents

1. Open the Document Information Extraction UI, as described in the tutorial: Use Trial to Set Up Account for Document Information Extraction and Go to Application or Use Free Tier to Set Up Account for Document Information Extraction and Go to Application.

   If you HAVE NOT just used the Set up account for Document Information Extraction booster to create a service instance for Document Information Extraction and subscribe to the Document Information Extraction UI, observe the following:

   • To access the Schema Configuration and Template features, ensure that you use the blocks_of_100 plan to create the service instance for Document Information Extraction Trial.

   

   • And make sure you’re assigned to the role collection: Document_Information_Extraction_UI_Templates_Admin_trial (or Document_Information_Extraction_UI_Templates_Admin if you’re using the free tier option). For more details on how to assign role collections, see step 2 in the tutorial: Use Trial to Subscribe to Document Information Extraction Trial UI, or step 3 in the tutorial: Use Free Tier to Subscribe to Document Information Extraction UI.

   

   • After assigning new role collections, Log Off from the UI application to see all features you’re now entitled to try out.

   

2. In the left navigation pane, click Schema Configuration.

   

Here, you find the SAP schemas. The Document Information Extraction UI includes preconfigured SAP schemas for the following standard document types: purchase order, payment advice, and invoice. In addition, there’s an SAP schema for custom documents (SAP_OCROnly_schema). You can’t delete or change SAP schemas. You can use them as they are, or create copies and adapt the list of fields according to your needs.

CAUTION:

When using the free tier option for Document Information Extraction or a trial account, be aware of the technical limits listed in Free Tier Option and Trial Account Technical Constraints.

To create your own schema, click Create and a dialog opens.

In the dialog, enter a name for your schema, purchase_order_schema, for example. Note that the name cannot include blanks. Further, select Purchase Order as your Document Type.

Click Create to create the schema.

Now, your schema shows up in the list. Access the schema by clicking on the row.

A schema defines a list of header fields and line item fields that represent the information you want to extract from a document.

Header fields represent information that is specific to your document and only occurs one time. This may include the document number, any sender information, or the total amount of the order. In contrast, line item fields represent the products that you ordered, where each line is one product, often with a certain quantity attached. Thus, the line item fields extract the information for each product in your order. Those may include the article number, the price and the quantity.

Document Information Extraction already includes a number of fields that it can extract. See here which header fields are supported and here which line item fields are supported. Additionally, you can define custom fields. In the next step, you’ll learn about both.

The image below shows an example purchase order. All the fields that you define in your schema in this tutorial are highlighted. The header fields represent all information outside of the table that occurs once. The line item fields represent all information within the table, which occurs per product. You can, of course, extend or reduce the information that you want to extract.

To define your first header field, click Add to the right of the headline Header Fields.

For each field, you have to enter a name, a data type, a setup type, and optionally a default extractor and a description. The available data types are string, number, date, discount, currency, and country/region.

The available setup types are auto and manual. The setup type auto supports extraction using the service’s machine learning models. You must specify a default extractor (standard fields supported by Document Information Extraction) for this setup type. It can only be used in schemas created for standard document types. The setup type manual supports extraction using a template. It’s available in schemas created for standard or custom document types.

As your first header field, add the number of your purchase order which identifies your document.

1. Enter an appropriate name for your field, purchaseOrderNumber, for example.

2. Select string for the Data Type. Note that a document number is a string, even though it consists of numbers, as it is an arbitrary combination of numbers without meaning. In contrast, price is an example of the data type number.

3. As all business documents have a unique identification, Document Information Extraction already includes a standard field. Select auto for the Setup Type and then select documentNumber for the Default Extractor.

4. Click Add to add the header field.

   

The field now displays in your list of header fields, where you again find all the information that you’ve just entered. You can edit or delete the field by clicking the respective icons on the right.

You’ve now added your first header field that uses a default extractor from Document Information Extraction. Next, you’ll add your first custom header field,

Click Add again to open the dialog.

1. Enter an appropriate name for your field, purchaseOrderStatus, for example.

2. Select string for the Data Type.

3. As Document Information Extraction offers no equivalent field, select manual for the Setup Type. Click Add to add the field.

   

You’ve now added your first custom field. Go ahead and add the list of header fields as shown in the table and image below. Pay attention to which fields have a default extractor and which don’t. Feel free to extend or reduce the list of header fields.

Next, you need to define the line item fields. As your first line item field, add the SKU (Stock Keeping Unit) that uniquely identifies an article.

Click Add to the right of the headline Line Item Fields.

In the dialog proceed as follows:

1. Enter an appropriate name for your field, skuNumber, for example.

2. Select string for the Data Type.

3. Select manual for the Setup Type and click Add to add the field.

The field now displays in your list of line item fields where you find all the information again that you’ve just entered.

You’ve now added your first line item field. Go ahead and add the list of line item fields as shown in the table and image below. Pay attention to which fields have a default extractor and which don’t. Feel free to extend or reduce the list of line item fields.

Once you’ve added all header and line item fields, the schema needs to be activated so that it can be used to extract information from documents. Right now, the schema has the status DRAFT, indicating that it cannot be used yet.

To activate the schema, click Activate.

Now, the status of your schema changes to ACTIVE. To make changes to your schema, you have to Deactivate it first.

Congratulations, you’ve created and activated your own schema for purchase order documents.

In the next tutorial: Create Template for Purchase Order Documents, you’ll create a template that uses your schema, and associate documents with your template to show the Document Information Extraction service where each field is located in the document.