As (anyone caring about data quality), I want that out of ordinary observations to be tagged automatically, so that
Data can be filtered easily
Data quality can be improved easily
As FIRI project admin, I want to utilize modern AI methods, so that project plan is fulfilled and we stay at the cutting edge.
Sub-cases:
1) As expert I want suspicious observations to be automatically marked to my checklist so I can go over them and do quality control.
2) As expert I want to be able to define and update what counts as a suspicious observation so that the get marked as needing check. (Request from Perhostutkijain Seura)
3) As data user, I want to filter observations based on automatic validation results, so that I get data that is suitable for my use.
4) As observer or data owner, I want to get notifications about possible problems in my observations, so that i can fix them. (Or tell that they are ok.
Requirements:
[ ] Works with large amounts of data, which cannot be checked manually
[ ] Does not slow down recording data with Vihko (trip form, mobile form, excel upload)
[ ] Works with data from third-party data entry tools (e.g. Pinnasieppo & Retkikasvio)
[ ] Works with all data in Laji.fi, including those copied from SYKE, Luke, MH ...
[ ] Does not require manual evaluation of all records (too much of them)
[ ] Does not mark records as erroneous without proper evaluation (could create circular argumentation)
[ ] Filters are created neutrally, and are not dependent on person
As (anyone caring about data quality), I want that out of ordinary observations to be tagged automatically, so that
As FIRI project admin, I want to utilize modern AI methods, so that project plan is fulfilled and we stay at the cutting edge.
Sub-cases:
1) As expert I want suspicious observations to be automatically marked to my checklist so I can go over them and do quality control.
2) As expert I want to be able to define and update what counts as a suspicious observation so that the get marked as needing check. (Request from Perhostutkijain Seura)
3) As data user, I want to filter observations based on automatic validation results, so that I get data that is suitable for my use.
4) As observer or data owner, I want to get notifications about possible problems in my observations, so that i can fix them. (Or tell that they are ok.
Requirements: