During server processing the tasks are carried out according to the following figure:

• All events occurring in a single task are executed in a single process.

• Different tasks can be distributed across different processes and therefore no shared information is available in different processes.

• Document Separation - If activated, document separation is performed as a preparation step of the batch. There are two separation algorithms available. One is based on image classification and is called Standard Document Separation. The other uses advanced algorithms for document separation and is called Trainable Document Separation. The separation events are placed in the project script sheet.

• Document Processing - the document processing can be divided into several substeps:

  Classification - Classification is the first step in document processing and can be treated independently from extraction although the extraction results can be used for classifying documents accordingly in a later step. All classification events are placed in the project script sheet.

  Extraction - The main purpose of this processing step is to recognize and extract relevant items on a document. There are different types of extraction events for either the document, the fields, or a script locator. All events related to the extraction events are placed in the corresponding class script.