Problem Management looks at incident trends.  When a trend is identified, it is called a "problem".  Thus, a problem typically represents multiple incidents.  If the problem is investigated further, once the root cause of the problem is known and a fix is created for it, it is known as a "known error". 

Problems and known errors are recorded as problem records and known error records in the Known Error Database. 

Every problem/known error has a category and a priority.  Problems that are not resolved may undergo escalation. 

Fixes and workarounds are also recorded in the Known Error Database.  Most fixes must be implemented in the IT infrastructure as a change.  To do this, the Problem Management process creates a change request, which is sent to the Change Management process. 

A decision must be made whether a fix will be created for a known error.  Not all known errors are worth fixing because of the cost involved or because of obsolete configuration items involved.

Problem Management also involves proactively looking for problems and resolving them.  This will reduce the number of future incidents.