DataFlux Data Management Studio 2.7: User Guide

Token Combination Rule Node

Add one or more of these nodes to a Token Combination Rules Group in order to generate one additional match code for each node, for each input string. You have the option of generating another match code for the default rule.

When enabled, each Token Combination Rule is applied if the tokens in the input string meet specified conditions. The conditions are categories and likelihoods that are applied to the token nodes.

Token combination rules move or remove one or more tokens in the default rule. You move tokens to account for input strings with tokens that are out of position. You remove tokens to account for input strings that lack tokens. The subset combination of tokens is combined with a sensitivity range to generate a match code.

The default rule consists of the ordered list of tokens, without actions, that is listed by default under Actions in the node's Properties dialog.

Used in:

• Match Definitions

Properties

Use this rule

When this check box is selected, properties become accessible and the rule will be used by the match definition.

Weight:

Optionally specifies a number from 1-100 that appears in the output for the rule, for use in resolving clusters.

Sensitivity range:

Specifies a range of sensitivity values that will be used to generate a match for token combination rule.

Conditions:

Lists all of the conditions that must be met before the specified actions are taken. If no conditions are specified, then the rule is always applied.

You can click icons to add, edit, and delete conditions. In the Condition dialog, you select a token, category, likelihood, and optional negation status. Negation represents a logical NOT, in which case the condition must not be met in order to take the specified actions.

Categories and likelihoods are made available for inclusion in conditions by adding Vocabularies to tokens in the match definition. For example, for the match definition EN Date (MDY), you could add a vocabulary that will tag the Day token with the category COULD_BE_MONTH and a likelihood of Medium. You could then apply that category and likelihood to a condition.

A condition is true if the token is tagged with the specified category and with a likelihood that is equal to or greater than the specified likelihood.

Actions:

Specifies how tokens will be combined when conditions are met. For each token, specify None to remove that token from the recombined output string. Specify a different token to indicate a rearrangement. Specify the same token to keep that token in the same position.

Output

Message

If any part of the input was found in the scheme, "Changes applied". Otherwise, the message will be, "No changes applied".

Result

A string of recombined tokens that is used to generate a match code.