1.6.3 - Quality dimension and Quality Rules
Description
Simpl shall provide the means for evaluating the quality of the self-descriptions based on at least the following quality dimensions of the FAIR-Principle (https://en.wikipedia.org/wiki/FAIR_data):
- Findability (F)
- Accessibility (A)
- Interoperability (I)
- Reusable (R)
It shall be possible to extend this list with further dimensions of quality rules. A good source is the ISO/IEC 25012 (https://iso25000.com/index.php/en/iso-25000-standards/iso-25012). For instance the quality dimensions:
- Data Completeness
- Data Consistency
For those quality dimensions Simpl shall provide quality rules that validate the quality for a given self-description. In contrast to the syntax (SIMPL-488) and semantic (SIMPL-490) validation the quality rules allow for more complex logical expressions, also combining fields. The checks go beyond semantic conformance, such as identifying inconsistencies between data instances, detecting redundant information, or verifying data integrity. The data quality rules can be classified into mandatory rules and recommended rules. Mandatory rules need to be fulfilled completely for all instances. Here are some example how those quality rules can look like it should be refined in the user story and during the implementation.
Dimension Completeness:
- Mandatory Rules: Those fields are required to be filled in (Unique Identifier, Title, Description, Location, Version, Creation date, Last update date, License, …)
- Recommended Rules: Those fields are recommended to be filled in (Type of Data, Keywords, Contact point, Schema, …) [weight: 5 for each field]
Dimension Consistency:
- Mandatory Rules: The Unique identifier needs to be unique across all self-descriptions. The creation date should be before the last update date.
- Recommended Rule: The location should be a valid URL. [weight: 5]
Dimension Findability:
- Recommended Rule: If at least 1 Keyword is set 5 Points, if at least 3 are set 10 Points, and for more than 5 Keywords 15 Points.
Dimension Interoperability:
- Recommended Rule: The type of data should be in the list of open formats provided by Simpl [weight: 5]
-Recommended Rule: The format under which the data is distributed should be in a list of common formats (e.g., xml) [weight: 5]
L2 - Detailed Requirement | Issue ID: SIMPL-1757 | Status: Proposed |
Please log in or sign up to comment.