com.wcohen.ss.expt

Class ExtractAbbreviations

java.lang.Object

com.wcohen.ss.expt.ExtractAbbreviations

public class ExtractAbbreviations
extends Object

Extracts abbreviation pairs (<short-form, long-form>) from text using an 'abbreviation distance metric' which evaluates the probability of a short-form string being an abbreviation/acronym of another long-form string. The probability is given by an HMM-based alignment between the two strings.

Sample command line:
java com.wcohen.ss.expt.ExtractAbbreviations ./train/abbvAlign_corpus.txt experiment_name

Citation: Dana Movshovitz-Attias and William Cohen, Alignment-HMM-based Extraction of Abbreviations from Biomedical Text, 2012, BioNLP in NAACL

Author:

Dana Movshovitz-Attias

See Also:

AbbreviationAlignment

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
class	ExtractAbbreviations.Stats

Field Summary

Fields

Modifier and Type	Field and Description
static String	SEPARATOR

Constructor Summary

Constructors

Constructor and Description
ExtractAbbreviations(String input, String output, String train, String gold)

Method Summary

Modifier and Type	Method and Description
protected void	addAbbreviationPairs(Map<String,Acronym> predictions)
protected AlignmentPredictionModel	loadPredictor()
static void	main(String[] args) Extracts abbreviation pairs from text. Usage: ExtractAbbreviations input experiment_name [gold-file] [train-dir]
protected void	mkdir(String dir)
protected String	outputAbbvs(Map<String,Acronym> predictions)
protected void	outputPairs(BufferedWriter bw)
protected ExtractAbbreviations.Stats	predictAndTest(int docID, List<String> corpus, List<Map<String,String>> trueLabels, BufferedWriter bw_abbvs)
protected void	predictAndTest(List<String> corpus, List<Map<String,String>> trueLabels)
void	run()
protected void	setTrainDir(String trainDir)

Methods inherited from class java.lang.Object

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

Field Detail

SEPARATOR

public static String SEPARATOR

Constructor Detail

ExtractAbbreviations

public ExtractAbbreviations(String input,
                            String output,
                            String train,
                            String gold)

Method Detail

run

public void run()
         throws IOException

Throws:

IOException

mkdir

protected void mkdir(String dir)

setTrainDir

protected void setTrainDir(String trainDir)

loadPredictor

protected AlignmentPredictionModel loadPredictor()

predictAndTest

protected void predictAndTest(List<String> corpus,
                              List<Map<String,String>> trueLabels)
                       throws IOException

Throws:

IOException

outputAbbvs

protected String outputAbbvs(Map<String,Acronym> predictions)

addAbbreviationPairs

protected void addAbbreviationPairs(Map<String,Acronym> predictions)

outputPairs

protected void outputPairs(BufferedWriter bw)
                    throws IOException

Throws:

IOException

predictAndTest

protected ExtractAbbreviations.Stats predictAndTest(int docID,
                                                    List<String> corpus,
                                                    List<Map<String,String>> trueLabels,
                                                    BufferedWriter bw_abbvs)
                                             throws IOException

Throws:

IOException

main

public static void main(String[] args)

Extracts abbreviation pairs from text.

Usage: ExtractAbbreviations input experiment_name [gold-file] [train-dir]