eus.ixa.ixa.pipe.nerc

Class Annotate

java.lang.Object

eus.ixa.ixa.pipe.nerc.Annotate

public class Annotate
extends Object

Annotation class for Named Entities in ixa-pipe-nerc.

Version:

2015-02-25

Author:

ragerri

Constructor Summary

Constructors

Constructor and Description
Annotate(Properties properties) It manages the use of the three different name finders: StatisticalNameFinder, DictionariesNameFinder and NumericNameFinder.

Method Summary

Modifier and Type	Method and Description
void	annotateNEs(ixa.kaflib.KAFDocument kaf) Classify Named Entities creating the entities layer in the KAFDocument using statistical models, post-processing and/or dictionaries only.
String	annotateNEsToCoNLL2002(ixa.kaflib.KAFDocument kaf) Output Conll2002 format.
String	annotateNEsToCoNLL2003(ixa.kaflib.KAFDocument kaf) Output Conll2003 format.
String	annotateNEsToKAF(ixa.kaflib.KAFDocument kaf) Output annotation as NAF.
String	annotateNEsToOpenNLP(ixa.kaflib.KAFDocument kaf) Output annotation in OpenNLP format.
String	convertToConLLTypes(String neType) Convert Entity class annotation to CoNLL formats.
StatisticalNameFinder	getStatisticalNameFinder() Get the statistical namefinder.

Methods inherited from class java.lang.Object

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

Constructor Detail

Annotate

public Annotate(Properties properties)
         throws IOException

It manages the use of the three different name finders: StatisticalNameFinder, DictionariesNameFinder and NumericNameFinder. In particular, if --dictTag option in CLI is off, statistical models are used (this is the default). If --dictTag is activated, it has two options, "tag" and "post": tag only tags with a gazetteer and "post" post-processes the probabilistic annotation giving priority to the gazetteer. Obviously, this option depends on the --dictPath parameter being correctly specified. The --lexer numeric option annotates numeric entities (dates, percentages, and so on) via rules. Finally, the oepc option post-process the output using the one entity per class hypothesis.

Parameters:

properties - the properties

Throws:

IOException - the io thrown

Method Detail

getStatisticalNameFinder

public StatisticalNameFinder getStatisticalNameFinder()

Get the statistical namefinder.

Returns:

the statistical namefinder

annotateNEs

public final void annotateNEs(ixa.kaflib.KAFDocument kaf)
                       throws IOException

Classify Named Entities creating the entities layer in the KAFDocument using statistical models, post-processing and/or dictionaries only.

Parameters:

kaf - the kaf document to be used for annotation

Throws:

IOException - throws exception if problems with the kaf document

annotateNEsToKAF

public final String annotateNEsToKAF(ixa.kaflib.KAFDocument kaf)

Output annotation as NAF.

Parameters:

kaf - the naf document

Returns:

the string containing the naf document

annotateNEsToOpenNLP

public final String annotateNEsToOpenNLP(ixa.kaflib.KAFDocument kaf)

Output annotation in OpenNLP format.

Parameters:

kaf - the naf document

Returns:

the string containing the annotated document

annotateNEsToCoNLL2003

public String annotateNEsToCoNLL2003(ixa.kaflib.KAFDocument kaf)

Output Conll2003 format.

Parameters:

kaf - the kaf document

Returns:

the annotated named entities in conll03 format

annotateNEsToCoNLL2002

public String annotateNEsToCoNLL2002(ixa.kaflib.KAFDocument kaf)

Output Conll2002 format.

Parameters:

kaf - the kaf document

Returns:

the annotated named entities in conll03 format

convertToConLLTypes

public String convertToConLLTypes(String neType)

Convert Entity class annotation to CoNLL formats.

Parameters:

neType - named entity class

Returns:

the converted string