medposttagger.java

dragontoolkit用于机器学习

package dragon.nlp.tool;

import dragon.nlp.*;
import gov.nih.nlm.nls.mps.POSTagger;
import dragon.util.*;
import java.util.*;

/**
 * <p>MedPost tagger adopted from gov.nih.nlm.nls.mps.POSTagger </p>
 * <p> </p>
 * <p>Copyright: Copyright (c) 2005</p>
 * <p>Company: IST, Drexel University</p>
 * @author Davis Zhou
 * @version 1.0
 */

public class MedPostTagger implements Tagger{
    private POSTagger tag;

    public MedPostTagger(){
        this(EnvVariable.getDragonHome()+"/nlpdata/tagger");
    }

    public MedPostTagger(String workDir) {
        try{
            if(!FileUtil.exist(workDir) && FileUtil.exist(EnvVariable.getDragonHome()+"/"+workDir))
                workDir=EnvVariable.getDragonHome()+"/"+workDir;
            tag = new POSTagger(workDir+"/lexDB.serial",workDir+"/ngramOne.serial");
            if(tag==null)
            {
                System.out.println("Failed to create POS tagger");
            }
        }
        catch(Exception e)
        {
            e.printStackTrace();
        }
    }

    public void tag(Sentence sent){
        List result, sentenceForTagger;
        String posLabel;
        Word word;
        int i;

        sentenceForTagger = new ArrayList();
        word=sent.getFirstWord();
        while (word != null) {
            sentenceForTagger.add(word.getContent());
            word = word.next;
        }
        result=(List)tag.tagSentence(sentenceForTagger);
        if(result.size()<=0) return;

        word=sent.getFirstWord();
        for(i=0;i<result.size();i++)
        {
            posLabel=(String)result.get(i);
            word.setPOS(posLabel,getPOSIndex(posLabel,word));
            word=word.next;
        }
    }

    protected int getPOSIndex(String pos, Word word){
        int posIndex;
        if (word.isPunctuation()){
            pos=word.getContent();
            posIndex=0;
        }

        else if (pos.startsWith("noun"))
            posIndex = Tagger.POS_NOUN;
        else if (pos.startsWith("verb"))
            posIndex = Tagger.POS_VERB;
        else if (pos.startsWith("adj"))
            posIndex = Tagger.POS_ADJECTIVE;
        else if (pos.startsWith("adv"))
            posIndex = Tagger.POS_ADVERB;
        else if (pos.startsWith("conj"))
            posIndex = Tagger.POS_CC;
        else if (pos.startsWith("det"))
            posIndex = Tagger.POS_DT;
        else if (pos.startsWith("pron"))
            posIndex = Tagger.POS_PRONOUN;
        else if (pos.startsWith("prep"))
            posIndex = Tagger.POS_IN;
        else if (pos.startsWith("num"))
            posIndex = Tagger.POS_NUM;
        else if (pos.startsWith("aux"))
            posIndex = Tagger.POS_VERB;
        else if (pos.startsWith("model"))
            posIndex = Tagger.POS_VERB;
        else if (pos.startsWith("compl"))
            posIndex = Tagger.POS_DT;
        else
            posIndex = 0;
        if(posIndex>0 && posIndex!=Tagger.POS_NOUN && word.isAllCapital()){
            posIndex=Tagger.POS_NOUN;
        }
        return posIndex;
    }
}