dragon.nlp.tool

Class Annie

java.lang.Object

dragon.nlp.tool.Annie

All Implemented Interfaces:

NER

public class Annie
extends Object
implements NER

Annie adopted from gate

Copyright: Copyright (c) 2005

Company: IST, Drexel University

Version:

1.0

Author:

Davis Zhou

Constructor Summary

Constructors

Constructor and Description
Annie()
Annie(String gateHome)

Method Summary

Modifier and Type	Method and Description
String	annotate(String original) The format of an annotated text is XML.
void	close()
ArrayList	extractEntities(String content) Extract entities from a given text.
ArrayList	extractFromDoc(Document doc) The return format is the same as the one by the method extractFromSentence.
ArrayList	extractFromSentence(Sentence sent) Difering from the method extractEntites, this method will save an entity in a term and thus all information such as entity type is kept.
void	initAnnie(String gateHome)
void	setAnnotationTypes(String[] arrType) Set the entity types of interest, for example, organization, location, person, etc.

Methods inherited from class java.lang.Object

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

Constructor Detail

Annie

public Annie()
      throws Exception

Throws:

Exception

Annie

public Annie(String gateHome)
      throws Exception

Throws:

Exception

Method Detail

initAnnie

public void initAnnie(String gateHome)
               throws Exception

Throws:

Exception

setAnnotationTypes

public void setAnnotationTypes(String[] arrType)

Description copied from interface: NER

Set the entity types of interest, for example, organization, location, person, etc. The interpretation of these types is subject to the implemented named entity recognizers.

Specified by:

setAnnotationTypes in interface NER

Parameters:

arrType - a set of wanted entity types.

close

public void close()

extractEntities

public ArrayList extractEntities(String content)

Description copied from interface: NER

Extract entities from a given text. Each extracted entity will be save in a Token. Duplicated entities will be merged. The frequency of a token is actually the occurrence frequency of the entity in the text. The type of the entity will be ignored.

Specified by:

extractEntities in interface NER

Parameters:

content - the text for processing

Returns:

a list of unique entities with their frequency.

extractFromDoc

public ArrayList extractFromDoc(Document doc)

Description copied from interface: NER

The return format is the same as the one by the method extractFromSentence.

Specified by:

extractFromDoc in interface NER

Parameters:

doc - a parsed document

Returns:

a list of extracted entities in their original order in the text.

extractFromSentence

public ArrayList extractFromSentence(Sentence sent)

Description copied from interface: NER

Difering from the method extractEntites, this method will save an entity in a term and thus all information such as entity type is kept. Moreover, duplicated entities will not be merged. Instead, all extracted entities will be returned in their orders in the text.

Specified by:

extractFromSentence in interface NER

Parameters:

sent - the sentence for extraction

Returns:

a list of extracted entities in their original order in the text.

annotate

public String annotate(String original)

Description copied from interface: NER

The format of an annotated text is XML. Each entity in the text will be tagged by a XML node and the attribute of the node will be the type of the entity.

Specified by:

annotate in interface NER

Parameters:

original - the text for annotation

Returns:

an annotated text