edu.stanford.nlp.parser.lexparser
Class Options

java.lang.Object
  edu.stanford.nlp.parser.lexparser.Options

All Implemented Interfaces:

Serializable

public class Options
extends Object
implements Serializable

Options to the parser which MUST be the SAME at both training and testing (parsing) time in order for the parser to work properly.

See Also:

Serialized Form

Nested Class Summary

static class

Options.LexOptions

Field Summary

boolean	coarseDistance Use coarser distance (4 bins) in dependency calculations
boolean	dcTags "double count" tags rewrites as word in PCFG and Dep parser.
boolean	directional Whether dependency grammar considers left/right direction.
boolean	distance Use distance bins in the dependency calculations
boolean	doDep Do a dependency parse of the sentence.
boolean	doPCFG Do a PCFG parse of the sentence.
boolean	flexiTag
boolean	forceCNF Forces parsing with strictly CNF grammar -- unary chains are converted to XP&YP symbols and back
boolean	freeDependencies if true, any child can be the head (seems rather bad!)
boolean	genStop
Options.LexOptions	lexOptions
boolean	nodePrune If true, inside the factored parser, remove any node from the final chosen tree which improves the PCFG score.
int	numStates
int	numTags
int	numWords
TreebankLangParserParams	tlpParams The treebank-specific parser parameters to use.

Constructor Summary

Options()

Options(TreebankLangParserParams tlpParams)

Method Summary

void	display()
TreebankLanguagePack	langpack() returns the treebank language pack for the treebank the parser is trained on.
void	readData(BufferedReader in) Populates data in this Options from the character stream given by the Reader r.
void	writeData(Writer w)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

numWords

public int numWords

numTags

public int numTags

numStates

public int numStates

lexOptions

public Options.LexOptions lexOptions

tlpParams

public TreebankLangParserParams tlpParams

The treebank-specific parser parameters to use.

forceCNF

public boolean forceCNF

Forces parsing with strictly CNF grammar -- unary chains are converted to XP&YP symbols and back

doPCFG

public boolean doPCFG

Do a PCFG parse of the sentence. If both variables are on, also do a combined parse of the sentence.

doDep

public boolean doDep

Do a dependency parse of the sentence.

freeDependencies

public boolean freeDependencies

if true, any child can be the head (seems rather bad!)

directional

public boolean directional

Whether dependency grammar considers left/right direction. Good.

genStop

public boolean genStop

distance

public boolean distance

Use distance bins in the dependency calculations

coarseDistance

public boolean coarseDistance

Use coarser distance (4 bins) in dependency calculations

flexiTag

public boolean flexiTag

dcTags

public boolean dcTags

"double count" tags rewrites as word in PCFG and Dep parser. Good for combined parsing only (it used to not kick in for PCFG parsing). This option is only used at Test time, but it is now in Options, so the correct choice for a grammar is recorded by a serialized parser.

nodePrune

public boolean nodePrune

If true, inside the factored parser, remove any node from the final chosen tree which improves the PCFG score. This was added as the dependency factor tends to encourage 'deep' trees.

Constructor Detail

Options

public Options()

Options

public Options(TreebankLangParserParams tlpParams)

Method Detail

langpack

public TreebankLanguagePack langpack()

returns the treebank language pack for the treebank the parser is trained on.

display

public void display()

writeData

public void writeData(Writer w)
               throws IOException

Throws:

IOException

readData

public void readData(BufferedReader in)
              throws IOException

Populates data in this Options from the character stream given by the Reader r.

Throws:

IOException