Rule Analysis is best done using a table or spreadsheet. Each rule makes one row in the table and each column is for an identity attribute. The cells in the table show the type of comparison that is required by the rule for that attribute.

Table 3 shows a table for 17 identity rules operating on 11 identity attributes labeled A1 to A11. The rules and attributes shown in Table, and the average number of candidates returned by various index schemes for these rules are based on tests performed on actual school enrollment information.

Although most attributes in this table require an "EXACT" match, some allow an approximate match by the Levenshtein Edit Distance function denoted by LED(0.xx) where the number 0.xx in parentheses indicates the matching threshold. For example, Rule 5 requires the Levenshtein similarity to be 55% or higher between the values of Attribute A3.

Back to OYSTER Reference Guide page

Click Prev The Alignment Process page

Click Next Index Design page