Initial State of Items

When items are processed by Duplicate Detect, they are processed in the context of a namespace definition. Several scenarios can occur depending on the way the namespace is defined, for example:

The application can auto-decision items just based on their key and secondary data fields. The application can also be set to have review turned off, meaning no review is needed before the unit of work is sent for processing by the next step of the payment processing system.
In this case, there are no potential duplicates because the unit of work for the items is processed automatically. As far as the system is concerned, if the keys and secondary keys of items match, a duplicate item has been discovered. The decisions made by the system are viewable and cannot be changed.
The application can have auto-decision rules, but review can be turned on.
In this case, potential duplicates are flagged and the unit of work is held for operator review of the potential duplicates before it is sent for processing by the next step of the payment processing system. Depending on the defined auto-decision rules, the number of potential duplicates may be limited because the application may stop flagging potential duplicates, if it has certain types of rules. For example, a rule can be written that auto-decisions everything to be a duplicate after 20 consecutive potential duplicates are processed.