Data quality monitoring and measurement in Reportnet Towards meaningful machine-readable metrics of data quality Hermann Peifer, EEA
Eionet Priority data flows 18 hand-made data flow reports during the period 1999 – ? 2016 ? 2016 ? 2016 ? ? ?
Magic ? Data quality Machine readable Meanigful metrics
Meaningful machine readable data quality metrics: 6 out of 12 = 50 % quality ???
Excursion: Mature Reportnet data flow using AutomaticQA with BLOCKER
DF1 and DF5: Report on all major roads, railways, … (1)
DF1 and DF5: Report on all major roads, railways, … (2)
Really Magic ? Data quality Machine readable Meanigful metrics Reportnet CDR changes in presentation layer feedbackStatus and feedbackMessage added END deliveries 1 Oct Oct 2015: Restart of AutomaticQA 250 envelopes delivered and released -30 envelopes (*.pdf, *.doc,*.shp only) -20 envelopes (very! old templates) envelopes: AutomaticQA restarted 133 WARNING: COULD BE wrong 122 ERROR: This IS wrong 432 INFO: No error found 1.Translate into 1-3 points 2.Calculate percentage 3.Bonus/Malus 4.Data quality (%) 5.Automatic Score Reportnet‘s Automatic Data checks ?