edu.stanford.nlp.parser.metrics
Class Evalb
java.lang.Object
edu.stanford.nlp.parser.metrics.AbstractEval
edu.stanford.nlp.parser.metrics.Evalb
- Direct Known Subclasses:
- Evalb.CBEval
public class Evalb
- extends AbstractEval
A Java re-implementation of the evalb bracket scoring metric (Collins, 1997) that accepts Unicode input.
"Collinization" should be performed on input trees prior to invoking the package programmatically.
A main method is provided that performs Collinization according to language specific settings.
This implementation was last validated against EVALB20080701 (http://nlp.cs.nyu.edu/evalb/)
by Spence Green on 22 Jan. 2010./**
- Author:
- Dan Klein, Spence Green
Constructor Summary |
Evalb(String str,
boolean runningAverages)
|
Evalb(String str,
boolean runningAverages,
boolean charLevel)
|
Methods inherited from class edu.stanford.nlp.parser.metrics.AbstractEval |
display, display, evaluate, evaluate, getEvalbF1, getEvalbF1Percent, getExact, getExactPercent, getLastF1, getNum, getSentAveF1, precision |
Methods inherited from class java.lang.Object |
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait |
Evalb
public Evalb(String str,
boolean runningAverages)
Evalb
public Evalb(String str,
boolean runningAverages,
boolean charLevel)
makeObjects
protected Set<Constituent> makeObjects(Tree tree)
- Specified by:
makeObjects
in class AbstractEval
evaluate
public void evaluate(Tree guess,
Tree gold,
PrintWriter pw)
- Overrides:
evaluate
in class AbstractEval
main
public static void main(String[] args)
- Run the scoring metric on guess/gold input. This method performs "Collinization." The default language is English.
The command line options are:
-arabic
Use settings for Arabic.
-y num
Specify the maximum yield of the guess trees.
-t
Evaluate bracketing at the character-level as proposed by Tsarfaty (2006).
- Parameters:
args
-
Stanford NLP Group