Arabic
To process invoices that are written in the Arabic language, use the following settings.
When a document in Arabic is scanned into the system using the Scan activity, select the AR language. This ensures that a recognition engine that is customized for the Arabic language is used for extraction
Next, ensure that the organization that generates Arabic invoices is configured as follows:
-
In the Extraction Settings, ensure that you add "AR" for Arabic to the Languages setting.
Lastly, select the Ignore Text PDF Layer setting on the General Settings.
It is important to select the option to ignore the PDF text layer because the built-in text layer does not contain any information about the right-to-left orientation of the Arabic language. When the text layer is ignored, the recognition engine is able to generate its own extraction results that honor the Arabic language direction.