Table of Contents
Match rules are used to identify good record pairs, to do this we need to define a series of potential rules that would be applied to each pair of records to determine if the pair is a high enough level to accept automatically or to create a task for a steward to review.
Each rule can be defined with a number of Concept Groups, each concept group is classed as mandatory or optional in the rule, in addition the weighting (ratio) between the concept groups in a rule can be defined.
- Add a new rule by clicking on the + dropdown in the Match rules level.
- Add a new concept group to a new rule, by clicking on the + dropdown in the Rule Purposes level.
Weightings (Accept Weight, Reject Weight and Mandatory)
Weighting ratios are the key to scoring with multiple concept groups, the general rule is defined as:
When there is a positive effect from the concept group (i.e. some similarity is detected) – the ratio of the score will be weighted according to the accept weighting of that concept group as a proportion of the total of all mandatory accept weighting(s).
In the event that there is no match for the concept group the reject weighting will be used.
Generally the accept weighting and the reject weighting can be the same (and the default reject weighting is the accept weighting if you miss it out), but if a concept group is very important and in the event it is significant to over value the lack of a match on the concept group then the reject weighting can be increased, similarly it can also be reduced, so a concept group only contributes to a match and not prevents one – however mandatory=false will also help you here.
|Acceptance Score||The high score to be used to indicate a good match that should be automatically accepted.|
|Review Score||The score that needs to be met to create a review task|
|Action text||Used with Auto and Task Reviewed linking it will be used to name the relationship between the records|
|Action to take when task resolved||Confusingly this is the action to take when the task is resolved either manually OR automatically|
|Currently Active||Is the rule active?|
|Find Duplicates in the same table||Prevents matching the same table data to itself, useful when each source is considered a golden source, and they should only be matched against each other.|
Click Save to save the Match rules.