MIT ID Data Problems

UPDATED: 4/9/1998 MIT ID Home Page

This document defines some of the data problems that have been identified with the deployment of MIT ID and proposes solutions for dealing with them going forward.

Duplicate ID Problem

Description: Two or more people with the same MIT ID.

Cause: This usually occurs when a search (manual or automated) incorrectly identifies a person and an existing MIT ID number is reused.

Resolution: All records associated with this MIT ID must be manually reviewed and the MIT ID modified where appropriate.

Proposed Solution: The MIT ID Database does not allow duplicate MIT ID's. If duplicate IDs exist across departmental systems only one of the records will appear in the MIT ID Database as determined by the feed from the Warehouse. The best time to identify Duplicate IDs is during the integration of multiple data sets in the Warehouse. Each case will need to be analyzed and cleaned up separately.

Multiple ID Problem

Description: An individual is assigned more than one MIT ID.

Cause: Usually assigned by different departments when there is inadequate searching of existing records prior to MIT ID assignment.

Resolution: Manually determine which MIT ID is "correct" and then change the ID for all records containing the "incorrect" ID(s).

Proposed Solution: It is proposed that a process be put into place where "incorrect" MIT IDs may be marked as obsolete and that they may be "linked" to the "correct" ID. Different algorithms will need to be employed to determine which ID should be the correct one. To prevent further multiple ID problems, an ID lookup operation must be performed before each individual is added to the MIT ID Database.