StandardWeightings

Select None to disable all built-in weighting presets. Only the custom rules defined in the local lexicon or included lexicons will be applied. This option is useful when you want complete control over all fuzzy match weighting rules and do not want to use any default or pre-configured settings.

When to Use

• You have highly specialized data or unique error patterns not covered by the standard presets.
• You want to ensure that only your explicitly defined rules are used for fuzzy matching.

The Full preset includes all standard numeric confusion rules (such as 'O' vs. '0', 'B' vs. '8'), as well as additional rules for common alphabetic swaps, punctuation, and visually similar characters. This preset is recommended for general-purpose data extraction where both numbers and letters may be misrecognized or mistyped.

Example Rules Included

• O0=0.25 (low cost for 'O' and '0' confusion)
• 3B=0.50, 5S=0.25, lI=0.25, Z2=0.50, etc.
• Additional swaps for punctuation, case, and similar shapes.

When to Use

• General OCR extraction where both numeric and alphabetic errors are common.
• Fields that may contain a mix of letters and numbers, such as serial numbers or codes.

The Numeric preset applies weightings that address common errors in numeric fields, such as digit swaps and confusion between numbers and similar-looking letters. It also includes rules for punctuation and control characters that may appear in numeric data.

Example Rules Included

• O0=0.25, I1=0.25, B8=0.25, S5=0.25, Delete(')=0.10
• Immutable=\t\r\n\f (prevents editing of control characters)

When to Use

• Fields that should contain only numbers, such as account numbers, IDs, or codes.
• Scenarios where numeric OCR errors are the primary concern.

The Currency preset extends the Numeric rules with additional weightings for currency symbols (such as '$' and 'S') and punctuation (such as '.', ',', and space). This preset is ideal for monetary values where OCR may confuse punctuation or currency symbols with digits.

Example Rules Included

• All rules from Numeric
• S$=0.20, s$=0.20, - .=0.30, .,=0.25, , .=0.25, etc.

When to Use

• Fields containing monetary values, such as totals, subtotals, or line item amounts.
• Scenarios where punctuation and currency symbol confusion is common.

The Labels preset includes all rules from Full and Numeric, plus extra weightings for label text, such as insertions and deletions of spaces and common label confusion (e.g., 'mn=0.5'). This preset is designed for extracting or matching field names, headers, or other label-like content where both alphabetic and numeric confusion may occur.

Example Rules Included

• All rules from Full and Numeric
• Insert( )=0.10, Delete( )=0.10, mn=0.5

When to Use

• Extraction or validation of field labels, headers, or section names.
• Scenarios where both character swaps and spacing errors are common in label text.