I2 (2 points)
https://knobit.org/id/task/rmd-match-quality/I2
Also as Turtle · JSON-LD · RDF/XML
| dcterms:license | https://creativecommons.org/licenses/by/4.0/ |
|---|---|
| owl:sameAs | https://bitbucket.org/reffa/ontology/id/task/rmd-match-quality/I2 |
| rdfs:label | I2 (2 points) en |
| schema:hasPart | rmd-match-quality I2: false-merge and false-split rates instead of merge counts |
| schema:hasPart | rmd-match-quality I2: the draft measured against the labelled sample |
| schema:identifier | I2 |
| schema:position | 5 |
| schema:text | Match quality, 1 each. (a) The pilot report counts merges, duplicate reduction and steward confirmations of the clerical queue, which say nothing about false merges: a finding reports it and quality_measure defines both a false-merge rate and a false-split rate, on a labelled sample, measured before go-live and periodically. (b) The draft measured on the labelled sample: the draft_on_sample band of all 12 pairs right, and draft_quality lists the false merges L-02 and L-03 (different people auto-merged) and the false splits L-04 and L-09 (the same person never compared, so no match); L-05, a true match in clerical review, may be listed as a split or not (checks.py). en |