Internally calculated data quality score examples - Data360_Govern - Latest

Data360 Govern Help

Product type
Software
Portfolio
Verify
Product family
Data360
Product
Data360 Govern
Precisely Data Integrity Suite > Govern
Version
Latest
Language
English
Product name
Data360 Govern
Title
Data360 Govern Help
Copyright
2024
First publish date
2014

The following is a selection of internally calculated data quality score examples.

Rules related

Three rules evaluate the technical asset column 1. The rule results were posted on 1/1/20.

Rules related

Configure the measure as follows:

  • Name: DQ Score using average.

  • Effective date: 1/1/20.

  • Weight: 100%.

  • Group Measure: No.

  • Rule Results Selection: Rules evaluate column.

  • Results Operation: Average.

    • Name: Column evaluated by rules.
    • Effective date: Today.
    • Weight: 100%.
    • Grouping measure: No.
    • Rule Result Selection: Column is evaluated by Rules.

      Two rules relate to column 1.

    • Results Operation: Average.
    • Rule Result Filters: None.
    • Conditions: None.

Score calculation = (.94 + .99 + .99) / 3 = 97.33, rounded to 97.3%.

Missing rule result

Three rules evaluate Technical Asset Column 1. The rule results were posted for two of the three rules on 1/22/2020.

Missing rule

Configure the measure as follows:

  • Name: DQ Score using average.

  • Effective date: 1/1/20.

  • Weight: 100%.

  • Group Measure: No.

  • Rule Results Selection: Rules evaluate column.

  • Results Operation: Average.

    • Name: Column evaluated by rules.
    • Effective date: Today.
    • Weight: 100%.
    • Grouping measure: No.
    • Rule Result Selection: Column is evaluated by Rules.

      Two rules relate to column 1.

    • Results Operation: Average.
    • Rule Result Filters: None.
    • Conditions: None.

Score calculation = (.94 + .99 + 0) / 3 = 64.33 rounded to 64.3%.

The UI displays that one result is missing from one rule.

Average of all rule results

The data quality score of the Columns technical asset type will be the average of all rule results. All data quality rules are of one type and no prior rule results exist.

  1. Create a 'Rule Type evaluates Column' relationship type. See Define relationships between assets and Relate quality rules to assets for more details.

  2. Relate columns to rules.

  3. Define a internally calculated Data Quality scoring definition, with the appropriate asset type.

  4. Define the following measure:

    • Name: Column evaluated by rules.
    • Effective date: Today.
    • Weight: 100%.
    • Grouping measure: No.
    • Rule Result Selection: Column is evaluated by Rules.

      Two rules relate to column 1.

    • Results Operation: Average.
    • Rule Result Filters: None.
    • Conditions: None.

Results

Score: 85% (.9+.8) /2, effective date: today.

Note: If rule results come in from three rules for one asset, but only two of the rules are actually related to the asset (with an evaluates predicate) then only two of those three rule results will be considered in the score.

If rule results are not posted for all the rules that the column is related to, the latest results for the missing rule will be used, until new results are available. For example, if tomorrow, new results are posted for rule 1, but not rule 2, rule 1 will, for example, be .85, but rule 2 remains at .80. The resulting score will be: 82.5% (.85 +.8)/2).

If the missing result for rule 2 becomes available later that day, and it is updated to .88, then the overall score will become: 86.5% (.85+.88) /2.

Give rule dimensions different weights

The data quality score of a Technical Asset Field will consist of 40% for the Accuracy Data Quality rule, and 60% for the Conformity Rule. Two measures are required.

  1. Create a 'Rule Type evaluates Field' relationship type. See Define relationships between assets and Relate quality rules to assets for more details.

  2. Relate fields to rules.

  3. Define a internally calculated Data Quality scoring definition, with the appropriate asset type.

  4. Define the following:

    Measure 1:

    • Name: Field evaluated by rules 1.
    • Effective date: 2/1/2021.
    • Weight: 40%.
    • Grouping measure: No.
    • Rule Result Selection: Field is evaluated by Rules.
    • Results Operation: Average.
    • Rule Result Filter: Rule Dimension = Accuracy.
    • Conditions: None.

    Measure 2:

    • Name: Field evaluated by rules 2.
    • Effective date: 2/1/2021.
    • Weight: 60%.
    • Grouping measure: No.
    • Rule Result Selection: Field is evaluated by Rules.
    • Results Operation: Average.
    • Rule Result Filter: Rule Dimension = Conformity.
    • Conditions: None.

Results

Score: 97.4% (.95 x .4) + (.99 x .6)

Measure 1

Measure effective date: 2/1/21 to current

Show Rule Results displays the following:

Rule 1, 2/1/21: .95

Measure 2

Measure effective date: 2/1/21 to current

Show Rule Results displays the following:

Rule 2, 2/1/21: .99