Index Rules (Bucketing)

Table of Contents

  1. Index Properties

Indexes are used for bucketing of records, this allows us to reduce the number of comparisons that is done between record, we use a bucket or index to group together “similar” records that will need to be compared before we decide on their likeliness.

To add an Index, under Data Modeling, go to the Index Rules option:

The drop down allows you to add a new index, or you can navigate to an existing one (shown is the default LabelSearch).

LabelSearch – is a default index that will always be defined by the system it groups together records based on their labels (which is defined by the table definition). It can be deleted but it will be recreated when a table changes (for that table).

Index Properties

The following is information on the index properties.

Concept GroupThe Defined Concept to use for this Index
Unique Identifier when same index name usedA name appendix to make it unique if index name is defined twice
Used for matching?Checkbox – does this index support matching?
Used for Searching?Checkbox – does this index support searching?
Maximum Cluster Size (performance) Maximum number of records to compare together based on this index
Exact Index – all the same values are considered to score 100%Checkbox – is this an exact index where any records with the same value are considered the same

Click save on completion and the Index Definition will be saved.