Linking Objects
To identify the key elements for onboarding, configure data onboarding rules and apply to Axon systems. After you enable automated onboarding from systems, the data sets and attributes are automatically created in Axon systems. You must also link the standard data quality rule in Axon to the rule in Informatica Data Quality. To automate data quality rules, choose to automatically create local data quality rules.
Linking Glossaries
In Enterprise Data Catalog, you can link a glossary to a field directly or through data domains. You can also assign recommended glossaries to a field in Enterprise Data Catalog. In Axon, you can link a glossary to a data domain discovery rule.
The onboarding criteria is based on the following associations between a field and glossary:
- Physical Field to Glossary
- In Enterprise Data Catalog, you can directly link a physical field to a glossary. Based on the physical field to glossary linking and the confidence score threshold that you configure in Axon, the data sets and attributes are automatically created and onboarded to Axon.
- For more information, refer the Add Business Title topic in the Informatica Enterprise Data Catalog User Guide.
- Data Domain to Glossary
- In Enterprise Data Catalog, a physical field can have a single or multiple data domains assigned to it. A data domain is a predefined or user-defined object that enables you to discover the functional meaning of field data, such as credit card number and Social Security number. You can link a data domain to a glossary. The physical field is indirectly linked to the glossary through data domains.
Based on the data domain to glossary linking and the confidence score threshold that you configure in Axon, the data sets and attributes are automatically created and onboarded to Axon.
- For more information, refer the Column and Field Assets and Data Domain Assets topics in the Informatica Enterprise Data Catalog User Guide.
- Physical Field to Recommended Glossary
- In Enterprise Data Catalog, when you link a physical field to a glossary, you can view the recommended glossary terms with the match percentage. Based on the confidence score for recommended glossaries in Enterprise Data Catalog and the confidence score threshold that you configure in Axon, the data sets and attributes are automatically created and onboarded to Axon. If there are more than one recommended glossaries, the attributes are created and onboarded based on the glossary with the highest percentage of recommendation and the confidence score threshold.
- For more information, refer the Add Business Title topic in the Informatica Enterprise Data Catalog User Guide.
- Glossary to Data Domain Discovery Rules
- In Axon, you can map a glossary to a data domain discovery rule. Enterprise Data Catalog uses the data domain discovery rules to discover matching field data or field name patterns from the metadata extracted by the resources. Enterprise Data Catalog discovers the attributes based on the association between the glossary and data domain discovery rules, and then onboards the discovered attributes to Axon.
- For more information, see Data Domain Discovery Rules.
- Default Glossary
- In Axon, you can specify a default glossary is specified in the Admin Panel. If a glossary association is selected in the onboarding rule for the system, Axon searches for the physical fields that are associated to a glossary. If a glossary association is not selected in the onboarding rule for the system, Axon uses the default glossary to onboard the objects from Enterprise Data Catalog.
Link Standard Data Quality Rule to a Rule in Informatica Data Quality
A standard data quality rule in Axon maps to a rule in Informatica Data Quality. You can map the standard rule to an existing rule in Informatica Data Quality, or you can automatically create a rule in Informatica Data Quality using CLAIRE recommendations.
Link Standard Rule to an Existing Rule in Informatica Data Quality
From Axon, you can link a standard data quality rule to an existing rule in Informatica Data Quality. After you link the rule, you can run the rule on Glossary object attributes that you onboarded from Enterprise Data Catalog.
The rule in Informatica Data Quality must conform to specific guidelines. For more information on the guidelines, see
Guidelines for Creating Rules in Informatica Data Quality.
Perform the following steps to link the rule:
1. From Unison search, open the Glossary object where you want to create a standard data quality rule.
2. Click the Data Quality tab.
3. In the Standard Data Quality Rules section, click the standard data quality rule that you want to link to a rule in Informatica Data Quality.
The standard rule opens.
4. Click Edit.
5. Click the icon next to Technical Rule Reference.
The Select Technical Rule Reference dialog box opens.
6. In the All tab of the Select Technical Rule Reference dialog box, select the rule from Informatica Data Quality. Select a project in the left pane, and then select the rule for the project in the middle pane. The right pane displays the number of input parameters for the option you select in the middle pane.
The following image shows a sample Select Technical Rule Reference dialog box:
7. After you select a rule from Informatica Data Quality, click OK to close the Select Technical Rule Reference dialog box.
8. Click Save & Close.
Guidelines for Creating Rules in Informatica Data Quality
Make sure that the rule in Informatica Data Quality conforms to the following guidelines:
- •The output value must be in the first column of the rule output.
- •The data type in the status output port of the rule must be of string type.
- •The output value for the rule must conform to the value criteria. The administrator can specify the output values that must be used by Axon to automate data quality rules. If the output values do not conform to the value criteria, Axon displays the data quality scores as 0% and marks all row counts as exceptions.
For example, if Axon requires the output value to be either Yes or No, but the data quality rule generates the output value as Y or N, Axon displays the Y scores as exceptions.
To specify the output values for the rules, see the Axon Data Governance 7.2.1 Administrator Guide.
If you have an existing rule in Informatica Data Quality that does not conform to the guidelines, modify the rule to meet the guidelines on the output criteria.
Automatically Create a Rule in Informatica Data Quality
If a data quality rule does not exist in Informatica Data Quality, you can automatically create the rule from the Axon interface. For a standard data quality rule in Axon, you can enter a description for the rule that you want to run on the Enterprise Data Catalog physical fields scanned. CLAIRE reads the description that you enter and intelligently recommends a single rule or multiple rules that can be run on the physical fields. If you agree to the recommendation, Axon creates the rule in Informatica Data Quality and then links the standard rule in Axon to the rule in Informatica Data Quality.
To recommend and create a rule automatically in Informatica Data Quality, the Axon Administrator must enable this option in the Admin Panel. For more information, see the Axon Data Governance 7.2.1 Administrator Guide.
1. From Unison search, open the Glossary object where you want to create a standard data quality rule.
2. Click the Data Quality tab.
3. In the Standard Data Quality Rules section, click the standard data quality rule that you want to link to a rule in Informatica Data Quality.
The standard rule opens.
4. Click Edit.
5. In the Technical Description field, start typing the rule description.
The CLAIRE icon
turns orange
.
6. Click the CLAIRE icon . The Select Technical Rule Reference dialog box opens.
7. In the Recommended tab of the Select Technical Rule Reference dialog box, review the recommended rule in the CLAIRE Recommendation and Rules sections.
The following image shows a sample Select Technical Rule Reference dialog box:
8. To modify the rule description, update the description in the Technical description field, and click the refresh icon 
. 9. Click OK when you are ready to use the recommended rule.
Axon creates a rule in Informatica Data Quality that conforms to the technical description that you entered.
10. Click Save & Close.
Guidelines for Entering Technical Descriptions
When you enter a technical description for a standard data quality rule, CLAIRE® reads and intelligently recommends one or more rules. Based on the description, CLAIRE® can provide up to a maximum of four recommendations to generate a rule. You can review the recommended rules and decide whether the rule meets your requirement. If you agree to the recommendation, Axon automatically creates the rule in Informatica Data Quality and associates the rule with the standard rule in Axon.
Consider the following guidelines when you enter a technical description so that CLAIRE® can recommend appropriate rules:
- •The description must be in English.
- •The description must be within 200 characters and 30 words.
- •The description must be in one sentence, and can contain letters, numbers, spaces, and special characters.
- •The following special characters are permitted: comma (,), hyphen (-), semi-colon (;), single quote ('), double quotes ("), angle brackets (< and >), equal sign (=), parentheses ({ and }), braces (( and )), square brackets ([ and ]), and period (.)
Examples of Empty Value Rules
If you want to check for null or empty values, you can use the keywords, such as "empty" or "null". You can also add lengthy and creative statements in the technical description.
The following examples show technical description of empty value rules that CLAIRE® can intelligently read:
- •Phone number should not be null.
- •Phone number must not be an empty string.
- •Phone number isn’t empty.
- •Phone number cannot be empty.
- •Phone number is expected not to be empty.
- •Phone number is required to be not empty.
- •Phone number should be an empty string.
- •Phone number is an empty string.
- •Phone Number is defined.
- •It's not acceptable for Phone Number to equal NULL.
- •Phone Number must be populated.
- •Phone Number should contain something.
When you check for a null or empty value, CLAIRE® suggests two rules in the following order:
Examples of Comparison Rules
You can enter a technical description creatively to specify a comparison. You can also enter compound sentences.
The following examples show technical description of comparison rules that CLAIRE® can intelligently read:
- •Down payment should be less than 10000.
- •Down payment is ten thousand.
- •Down payment should be greater or equal to 10000.
- •Down payment cannot be less than ten thousand.
- •Down payment shouldn’t be greater or equal to 10000.
- •Down payment must be smaller than 10000.
- •Water depth holds a number greater or equal to zero.
- •Water depth is going to either be 2 or be above 2.
- •Value cannot be less than 1,000.
CLAIRE® suggests a single rule when you check for a comparison.
Examples of String Length Rules
You can enter length-related statements in different ways in a technical description. You can use the words, such as "characters" or "digits", in the technical description.
The following examples show technical description of length rules that CLAIRE® can intelligently read:
- •Product Code must not be longer than 9 characters.
- •Product Code must not be longer than 9 symbols.
- •Product Code mustn't be longer than 9 letters and digits.
- •Product Code is 9 characters long.
- •Product Code is 10 digits.
- •Product Code must be less than or equal to nine characters.
- •Product Code shouldn't be greater than 9 characters.
- •Product Code must be shorter than nine characters.
- •The length of Policy Description should have value other than 10.
- •ID's length equals 0 or is above 0.
- •Policy Description has a length that's more than 10.
- •Policy Description is at least 10 long.
- •ORG ID has at least one character.
CLAIRE® suggests a single rule when you check for a string length.
Examples of Data Type Rules
You can enter synonyms and lengthy sentences or keywords, such as "number" or "date", in the technical description.
The following examples show technical descriptions of data type rules that CLAIRE® can intelligently read:
- •Expiration Date must be a date.
- •Expiration Date must be of type date.
- •Lease number must be of type integer.
- •Last Name isn’t a number.
- •Employee ID is numeric.
- •Expiration Date can have values of type date only.
- •Expiration Date should contain the value of date.
- •It can only have values of type date.
When you check for a data type value, CLAIRE® suggests a single rule based on whether the data type is a date or number.
Examples of Range Rules
In a technical description, you can enter a markup that explicitly denotes the range and inclusion of boundary values. You can also enter complex language, including some compound sentences, in the technical description.
The following examples show technical description of range rules that CLAIRE® can intelligently read:
- •Diameter should be within [1; 2]
- •Diameter should be in (1; 2)
- •Diameter is in [1.5; 2.5)
- •Diameter shouldn't be in [1 ; 2)
- •Diameter exceeds 0 and doesn't exceed 5.
- •Diameter is bigger than zero and also diameter is smaller than five.
- •Diameter is between one and five, but it cannot be one or five.
- •Diameter is at least 1 or higher, but not as high as 10.
- •Diameter can be equal to 1, or be between 1 and 10.
When you check for a range, CLAIRE® suggests recommendations based on the following scenarios:
- •If you enter brackets, such as "X must be in (1;2]", to check for a range, CLAIRE® suggests a single rule because you have explicitly specified whether the boundary values have to be included or not.
- •If you check for a range in the enhanced format, such as "X must be between 1 and 2", CLAIRE® suggests four recommendations, one for each combination of inclusion and exclusion of boundary values. Due to language ambiguity, CLAIRE® cannot interpret if you want to include or exclude the boundary values.
Examples of List Rules
In a technical description to define a list, you can enter markups or delimiters, such as curly braces ({}) and semicolon (;). To facilitate naturally written sentences, you can include conjunctions in the delimiters. Enclose each item of a list within double quotes ("").
The following examples show technical description of list rules that CLAIRE® can intelligently read:
- •Index must be {07; 10; 14; UU}
- •Index is {07;10;14;UU}
- •Index is one of {07 ; 10 ; 14 ; UU}
- •Index is one of {07; 10; 14; UU;}
- •Index is one of {07; 07; 10; 14; UU}
- •Index is one of {07,10,14,UU}
- •The value of ARM Index is "0", "1" or "2"
- •ARM Index shall be equivalent to "yes", "no" or "na"
- •ARM Index will be one of these: "yes", "no", "na"
- •ARM Index should have values within the limit of "yes", "no" and "na"
- •"Yes", "no" or "na" can only be the values that ARM Index can hold.
CLAIRE® suggests a single rule when you check for a list.
Note: Use the double quotes to indicate lists. If at least one double quote is found in the sentence, CLAIRE® attempts to read the list items.
Limitations
The following types of sentences are not supported in a technical description:
- •You cannot combine different types of checks. Ranges and length check sentences are not compatible. For example, sentences such as "Length of X must be between 1 and 10" are not supported. You can either perform a comparison-type length check, such as "Length of X must be below 1", or perform a range check on the value of the glossary term, such as "X must be between 1 and 10", but not both.
- •Compound statements such as "X must be not null and be a number" are not supported.
- •Axon does not perform a validity check for ranges. For example, if you enter "X must be below 9 and above 10", there are no values that satisfy this range check and there are no validity checks performed. CLAIRE® detects the entered range, extracts two numbers, sorts them, and forms a range check so that the value is between those two numbers.
- •If you enter the description without the brackets, range check can only address a single range. A compound sentence for a range is not supported if you want to check whether a value is outside a particular range. For example, the following sentences are not supported:
- - X to be below nine or above 10.
- - X is lesser than ten or also x is greater than or equal to two hundred.
- •Axon performs a proper range check based on the description that you enter. For example, both the description "X must be in [1; 2]" and "X must be between one and two" generate the proper rules. If the value is outside the range, for example, "X must not be between one and two", Axon does not generate the proper rule. Instead, for example, "X must not be between [1;2]" generates the proper rule.
Create Local Data Quality Rules Automatically
After you link a standard data quality rule in Axon to a rule in Informatica Data Quality, you can choose to automatically create local data quality rules for the Glossary object attributes. Make sure that each Axon attribute is linked to only one column in Enterprise Data Catalog. In Axon, the Axon Status field for the standard rule must have the Active value
Axon creates local data quality rules when certain conditions are met. To know the conditions, refer to
Conditions for Automatically Generating Local Data Quality Rules.
Perform the following steps to create local data quality rules automatically:
1. From Unison search, open the Glossary object that is associated with the standard data quality rule.
2. Click the Data Quality tab.
3. In the Standard Data Quality Rules section, click the standard data quality rule for which you want to create local data quality rules.
The standard rule opens.
4. Click Edit.
5. Set the value of the Axon Status field to Active.
6. Select the Create Local Rules Automatically check box.
7. If you selected a rule that requires multiple input parameters, in the Input to Glossary Mapping section, associate each input parameter to a Glossary object.
Note: Make sure that one of the Glossary objects that you map to an input parameter is the Glossary object that is linked to the attribute.
The following image shows a sample Input to Glossary Mapping section:
8. Click Save & Close.
