Überspringen zu Hauptinhalt

New inter-judges agreement having hours analyzed by one or two evaluators was 80% and for instances analyzed by the three evaluators try 76%

New inter-judges agreement having hours analyzed by one or two evaluators was 80% and for instances analyzed by the three evaluators try 76%

We determined inter-evaluator arrangement since a simple proportion between the instances when all the newest evaluators gave the same consequences (all proper or the completely wrong) divided because of the final number away from times evaluated by same evaluators.

Error study

We assessed 100 program errors chosen at random regarding predications considered just like the not true professionals by the around three judges. From this study it turned into clear one to completely wrong answers returned because of the the system was indeed largely because of incorrect predications acquiesced by SemRep from the source data files (MEDLINE citations).

Mehr Lesen
An den Anfang scrollen