DataFinder presets common data warehousing verification rules for you. When the event warehousing does not meet the verification rules, the corresponding event will be intercepted. At the same time, DataFinder also supports you to customize data verification rules according to business needs. This article introduces the configuration points for creating custom verification rules.

• Currently only privatization v4.4.1 and above is supported.
• The current function is turned off by default. If you wish to use the ability to customize validation rules, you can contact BytePlus technical support to turn the function on.
• For details on DataFinder's preset validation rules see Reference: Preset Data Validation Rules, Before creating a custom validation rule, you can check whether the preset validation rule meets your business requirements.

1. After logging in and entering the DataFinder console, click Data Management > Data Governance > Validation Rule Configuration in the top navigation bar. After entering the Validation Rule Configuration page, click the New Validation Rule button in the upper right corner.
2. Configure custom validation rule parameters in the pop-up page.
   

   1. Configuration basic information.

      Parameter	Configuration instructions
      rule name	Custom rule name for subsequent identification of custom validation rules.
      rule type	Currently supports event quantity, event attribute validation, user attribute validation three types of rules. Event Volume: A validation rule that verifies the number of reported events over a period of time. Event attribute verification, user attribute verification: Verify whether the event/user attribute meets the verification rules.
      Validation Object & Validation Rules	Select the validation object according to the interface prompts and configure the validation rules. The validation rules of the event attribute need to specify the event before selecting the event attribute. Details of event and attribute validation rules can be found in the following Reference: Validation rule descriptionsections. The verification rule supports adding multiple, and the logical relationship between multiple verification rules is "OR", that is, if one of the verification rules is satisfied, the alarm and data interception can be carried out according to the alarm rule. Support configuring filter conditions for each check rule configured, and the logical relationship between multiple filter conditions is "AND".

   2. Configure alarm settings.
      You can choose whether to turn on the monitoring alarm of the verification rule as needed. After opening, when the subsequent data is put into storage, if the abnormal data that does not meet the verification rule reaches a certain amount, an alarm message will be triggered according to the alarm rule.

3. Data interception settings. You can choose whether to intercept abnormal data as needed. After opening, when subsequent data is put into storage, abnormal data that does not meet the verification rules will be intercepted according to the interception rules and will not be put into storage.
   

   tip

   Event property/user property rules need to be configured, event volume rules do not need to be configured.

After the custom verification rule is created, you can view the list of custom verification rules on the verification rule page, and support operations such as opening and closing the verification rule.

• After closing the validation rule, the validation rule will no longer take effect in the future. If you no longer wish to perform the relevant warehousing validation, you can turn off the corresponding validation rule.

  warning

  • If you configure the alarm notification after the verification rule is triggered, after a verification rule is closed on the verification rule page, after the subsequent verification rule is triggered, the data will not be intercepted and put into storage, and can be put into storage normally; but the alarm message may also be triggered and sent.
  • If you also want to turn off alarm notifications, you need to click the Edit button to turn off alarms in the Monitor alarm configuration of the validation rule.
    
  • The verification rule of the event volume type does not support closing the alarm. If you do not want to trigger the verification into the warehouse or receive the alarm notification in the future, you can directly delete the corresponding verification rule.
• Closed validation rules support direct deletion.
• The created verification rule can be edited and modified. Editing and modifying only supports modifying the alarm and interception settings of the verification rule, and the verification rule does not support modification.
• You can also check the number of alarms and abnormal data of the corresponding verification rule on the verification rule list page, and click the corresponding data to view the details.

Event Volume Validation Rule Description #

Event volume verification is used to monitor the amount of reported data for a specific event tracking, and to trigger an alarm when the amount of event reports is greater than or less than a certain number within a specific time range.
For example, an alarm will be raised when the order payment event exceeds 10,000 within 5 minutes.
The time interval for verification can be divided into minutes, hours, and natural days, with a minimum interval of 10 minutes.

Attribute validation rule description #

Event/User Properties The following validation rules can be selected on demand according to the interface prompt.

Q1 What situations do validation rules apply to? #

Verification rules are used to control data quality and ensure that the content of data reports after event tracking is online continues to meet the design specifications.
For example, if the value of the "whether or not a member" attribute must be passed, and the reported content must be 0 or 1, then the data that has not been reported as "whether or not a member" attribute or the reported content is incorrect can be intercepted and alerted through rule verification, and the researchers will be notified to correct it in time.

Q2 What is the difference between the data verification alarm and the Kanban monitoring function? #

The data verification alarm is a monitoring alarm for the original reported data, which can only monitor the quality of the data and cannot monitor the indicators (for example: click through rate).
Kanban monitoring is the monitoring and alerting of the data built into the warehouse (ie: the data intercepted after processing), which is suitable for business indicator monitoring.