Replace Quality Scores with Categories
See original GitHub issuePrerequisites
- [ X ] Put an X between the brackets on this line if you have done all of the following:
- Running the latest version of Constellation
- Attached the Support Package via
Help
>Support Package
- Checked the FAQs: https://github.com/constellation-app/constellation/wiki/FAQ
- Checked that your issue isn’t already filed: https://github.com/constellation-app/constellation/issues
- Checked that there is not already a module that provides the described functionality: https://github.com/constellation-app/constellation/wiki/Catalogue-of-Repositories
Description
The Quality Control View is currently designed to run ‘rules’ across data in a graph in order to identify data quality issues, and to provide feedback on these issues in the form of a quality ‘score’ (0 to 100 - with 100 representing the worst quality) and a ‘reason’ for that score. Since these scores are essentially arbitrary (what does a score of 70 actually mean?), I propose we instead move to a system where we replace data quality scores with data quality categories, which I believe would be more interpretable. I imagine these categories might work much like a logging system, where there is a hierarchy of categories representing the severity or impact of the data quality issue (ie. FATAL, ERROR, WARNING, INFO). We could then assign each rule to a category - so the ‘unknown type’ rule might map to the ‘INFO’ category, whereas the ‘blocked by compliance’ rule would map to the ‘FATAL’ category. You might also add the ability to configure the rank of each category, so a user could specify which categories are important to them.
Issue Analytics
- State:
- Created 3 years ago
- Comments:8 (4 by maintainers)
Top GitHub Comments
In terms of UI - I’d like to get your thoughts on something like this @arcturus2 . It’s just a Paint mock up. Pre-defined categories, which can be filtered to show only a certain category. EDIT: Realized I replaced
Comment
instead ofScore
.Good work @aldebaran30701. I tested this and looks good. Few minor changes may be;
Giving a label to the preference json (see below) like “customRules=” similar to the recentValues.
={“Missing type”:“WARNING”,“Unknown type”:“DEFAULT”,“Identifier inconsistent with type”:“INFO”}
The alert box w/o the rule(see attached) can display a message like “No rules exist for this Identifier”