How to work with Semaphore label settings
- Last Updated: May 13, 2026
- 8 minute read
- Semaphore
- Documentation
Explanation
Knowledge Model Management (KMM) offers the ability in the model to override at the concept label level (preferred or alternative labels) certain behaviors that are defined in the rulebase templates.
Note: These settings do not apply to the entire concept. You can only apply the settings on each individual label in a language. (For example, if a concept has a copy of the preferred label in four different languages, you will have to adjust the settings for each individual label to have them apply for all of the preferred labels.)
This article explains each label setting and how it affects publishing.
Assumptions
- You have a model
- You have the MANAGER, EDITOR, or CONTRIBUTOR role for the model.
- You have a working classification strategy and implementation for your model.
Method
To access label settings for a concept:
- Open a model (in this example, Space Missions), and click Open Master or open a task.
- Navigate to a concept.
- Hover on the preferred label or any alternative label. Click the gear icon that appears.
- Adjust settings in the Edit Label Settings window (explained below), and then click Save.
Label Settings Explanations
Case Sensitivity
Description: Determine if a label must consider letter case when matching for classification.
Options:
- On - A match will occur when the word in the content is exactly as it is in the Label. (For the preferred label Apple, the setting will override all case properties in the templates and force matches to only happen when the content mentions “Apple”. It will not match on “apple”.)
- Default (selected by default) - Whatever case settings exist in the rulebase template will be honored.
- Off - A match will occur when the word in the content matches the letters of the label in any case. (For the preferred label Apple, the setting will override all case properties in the templates and force matches to happen when the content mentions “Apple” or “apple”.)
Rulebase Actions
From Semaphore 5.12 forward, this attribute will move from the label onto the Concept or Concept Scheme, so it will not appear on the label settings panel.
Description: As applied to a preferred label, will determine if a rulebase will be created for a concept. As applied to an alternative label, will determine if individual rules will be created for that alternative label. (Rulebases are used to tag content during classification.)
As a result, options set for preferred labels will take precedence over those set for alternative labels. If the preferred label is set to not generate a rulebase, then no settings for alternative labels will apply, since no rulebase at all will exist.
- Do not generate rulebase - The concept will not have a rulebase created, and therefore cannot be returned by CLS.
- Generate rulebase (default) - The concept will have a rulebase created, and therefore can be returned by CLS.
- Substitute in manually edited rulebase - Do not use the rulebase template to create the rulebase for the concept. Instead, use the rulebase found in the specified static rulebases directory. (See The Kid Writer for how to implement a manually edited rulebase.)
Options (for alternative labels):
- Do not generate rulebase - The alternative label will not have rules created for it, and they therefore will not be returned by CLS as part of the rulebase.
- Generate rulebase (default) - The alternative label will have rules created for it, and they therefore will be returned by CLS as part of the rulebase.
- Substitute in manually edited rulebase - Setting this on an alternative label will have no effect on rulebase generation.
Behavior in rulebase
Description: This value is used in templates as a rule filter which determines whether or not any given rule is generated for the label. The value is used in the behaviourtypes attribute in the templates (see behaviourtypes for details).
Options:
- Allow phrase variants (default) - This will generate rules for the label where the template attribute, behaviourtypes is specified as “Allow phrase variants (default)” OR the attribute is not specified on the rule.
- Treat as exact phrase only - This will generate rules for the label where the template attribute, behaviourtypes is specified as “Treat as exact phrase only” OR the attribute is not specified on the rule. See behaviourtypes#treat_as_exact_phrase_only for detailed instructions on how to set up your templates if you want to use this option.
Note: these are the out of the box options for this setting. Additional values can be added by the user to be used in customer specific classification strategies. Any additional values are used in the same manner, i.e. as a filter on a template rule.
Influence in rulebase
Description: Change the weight (or score) that a label receives in the published rulebases. The setting overrides the weights in the templates by multiplying them by set amounts.
(Example: You may want to classify a document as being about the company Apple, and it might refer to “Apple” or “Apple, Inc.,” and its products as “Apple computers.” Since apple is a common word, it becomes important to clarify when “apple” should help classify a document as being about the company. This is easy to address in KMM by adding the alternative label “apple pie” and setting the Influence in Rulebase to “none.” Publisher will create a rule that tells Classification and Language Service (CLS) to ignore instances of apple that it sees in the phrase “apple pie”.)
Options:
- High - The label will receive double the score that is assigned in the rulebase template.
- Low - The label will receive half the score that is assigned in the rulebase template.
- None - The label will receive a zero score (it will never match in classification).
- Normal (default) - Uses the scores assigned in the rulebase templates.
- Tag if present - The label will receive a score of 100 regardless of the score assigned in the rulebase templates.
The values used to manipulate the ‘weight’ attribute in a rule (e.g. “High” means ‘double’, “Low” means ‘halve’, etc.) can be modified if needed. See the “Rulebase Influence Handler” on this page, How to work with Semaphore label settings.
Stemming
Description: Determine if a match occurs based on the stem of a label matching the stem of a word in the content (e.g., plurals or verb conjugations).
Options:
- On - A match occurs where the stem of the preferred label matches the stem of words that occur in the content. (For example, the label “bake” will match with “bake”, “bakes”, and “baked” in the content.)
- Default (selected by default) - Whatever stemming settings exist in the rulebase template will be honored.
- Off - No match occurs between the stem of a label and stems of words in the content.
Used for concept mapping
Description: Determine whether a label is used in the concept mapping output of the Semantic Enhancement Server (SES). (Concept mapping works as an integrated service provided by SES. Concept mapping occurs when a user enters a search term into a search box and is provided with matching concepts from the model.)
Options:
- On - The label is available to use in concept mapping.
- Default (selected by default) - Whatever settings exist in the global templates will be honored.
- Off - The label is not available to use in concept mapping.
Escape special characters (advanced setting)
Description: Determine whether Semaphore special characters in the label (i.e., wildcard characters like * or ?) are “escaped” to prevent Classification Server treating them a special characters. (See data for a list and description of wildcard characters.)
Options:
- On - Special characters are escaped for handling. For the label door #2, the hash mark will be treated as a literal hash mark (#).
- Default (selected by default) - Whatever settings exist in the publishing configurations will be honored. (The out of the box global setting is Off.)
- Off - Special characters are not escaped and receive special handling. For the label door #2, the hash mark will be treated as a wildcard character that means “zero or more numeric digits”.
A-Z entry (advanced setting)
Description: The A-Z entry setting sets a label to be evaluated when using the SIS “az” command (How to work with Semaphore label settings). This command is typically used when creating an A-Z navigational UI component in downstream search and navigation systems. Semaphore includes an example of such a widget in its integration examples.
The default setting for A-Z entry is set at a global level in the <alphabeticalHandler> bean found in the ModelInterface.xml file in the Publisher \import directory. Out of the box, the default level is set to ‘false’.
Options:
- On - The label is available to display in the SES A-Z list.
- Default (selected by default) - Whatever settings exist in the rulebase template will be honored.
- Off - The label is not available to display in the SES A-Z list.
Use for search-as-you-type (advanced setting)
Description: Determine whether a label is used in search boxes integrated to the Semantic Enhancement Server (SES). (This feature occurs when a user begins entering enters a search term into a search box. As the user types the word, search-as-you-type (SAYT) suggestions for possible concepts in the model appear for the user to select.)
Options:
- On - The label is available for SAYT in SES.
- Default (selected by default) - Whatever settings exist in the rulebase template will be honored.
- Off - The label is not available for SAYT in SES.
Outcome
Changes you make to the settings will be implemented the next time you publish the model.