<< Click to Display Table of Contents >> Navigation:  Modules > Workflow Automation > Decision Trees

A Decision Tree is used to run records of a target table through a battery of tests (usually to test for data integrity), finally converging on a single outcome and remediation advice (if required) and/or action.

A Decision Tree is comprised of one or more nodes.  Each node, when evaluated, will result in either a true or false outcome.  Depending on the node's outcome, the tree will branch to a new node, and continue testing (or isolating the root cause of the previous test.)

Branching through the tree's nodes will continue, until a Terminal Node is reached.  The Terminal Node contains remediation advice (placed there by the tree's designer).  The remediation advice is copied over to the target record for subsequent reporting.

Decision Trees are integrated with the Workflow engine.  Any given node may fire a workflow, which can then be used to perform notification, solicit input, or even directly perform remediation.

Each time a Decision Tree is run against a target record, a Thread is created.  The history log in the thread shows how the tree progressed through the various nodes.

Decision Trees are managed by the Housekeeping function, which generally runs every five minutes.  Since a Decision Tree may take an appreciable amount of time to run, the MIE enforces a maximum number of records that will be submitted to the tree at any given time.  Records are submitted oldest first ("oldest" defined as the longest time since the record was last submitted to the tree).