package edu.stanford.nlp.semparse.open.dataset; import java.util.List; import edu.stanford.nlp.semparse.open.dataset.entity.TargetEntity; public interface Criteria { public List<TargetEntity> getTargetEntities(); /** Number of criteria **/ public int numCriteria(); /** Number of matched criteria **/ public int countMatchedCriteria(List<String> predictedEntities); /** Return a custom IR score */ public IRScore getIRScore(List<String> predictedEntities); /** * Return the correctness score. Between correct candidates, the one with * higher correctness score is more correct. * * Normally, this is just getIRScore().f1 */ public double getCorrectnessScore(List<String> predictedEntities); }