The link index file is a file that is generated by OYSTER for every OYSTER run. This file defines the clusters by assigning each input reference a link value so that references found by OYSTER to identify the same real-world entity belong to same cluster. This file contains three pieces of information for each reference processed during the OYSTER RUN.

• RefID - Identifies the reference by combining the source name and the reference ID of the reference.
  • The source name is defined by the user through the Name attribute of the OysterSourceDescriptor tag discussed in section "3.2 <OysterSourceDescriptor> Tag" of this document.
  • The reference ID of the reference is defined by the user through assigning the @RefID keyword to an Attribute attribute of an Item tag discussed in section 3.5 <Item> Tag of this document.
• OysterID - this is the unique link value defined by OYSTER and is used to identity references that belong to the same cluster.
  • Each cluster of references define the same real-world identity
  • The same OysterID is assigned to each member of a cluster.
• Rule - This identifies which rules were used when matching this reference to other references.
  • Rules are assigned unique identifiers by the user through the Ident attribute of the Rule tag discussed in section 2.5 <Rule> Tag of this document.
  • A record may have more than one rule listed if they were matched on more than one rule.
  • A rule of "null" will be listed if no matches were found for the record.

An example Link Index is illustrated in Figure 9. This link Index defines three clusters, {source1.1, source1.3, source1.5}, {source1.2, source1.4}, and {source1.6} with cluster identifier (link) values of QN5Y16P5HK5JX6FA, ZKCZ2FFW4R6RICUV, and E255F96M3COGM323 respectively.

Back to OYSTER Reference Guide page

Click Prev Identities Tag page

Click Next Identity Change Report page