Package gannuNLP.dictionaries

Examples of gannuNLP.dictionaries.Wiki

31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
  {   
    super(dict,name);
    this.dict=dict;
    File tmp=new File("./data/matrices/"+dict.getCompleteName().replace(">", "@@@@@")+"/");
    tmp.mkdirs();
    Wiki wiki=(Wiki)this.dict;
    for(Input doc:targets)
    {
      System.out.println("Calculating matrices for "+doc.toString());
      for(AmbiguousWord word:doc.getAmbiguousWords())
      {       
        if(word.getCurrentLemmaObject()!=null&&word.getSenses().size()>0&&word.getSenses().size()<150)
        {
          System.out.print(".");
          File fout=new File("./data/matrices/"+dict.getCompleteName().replace(">", "@@@@@")+"/"+Dictionary.normalizeLemmaforFile(word.getLemma())+".gmx");;
          if(!fout.exists())
          {           
            SmallPair p=new SmallPair(word.getLemmaObject());           
            for(int i=0;i<p.getLemma().getSenses().size();i++)
            {
              p.getCounts()[i][i]=(int)wiki.getCounts(p.getLemma().getSenses().get(i),p.getLemma().getSenses().get(i));
              for(int j=0;j<p.getLemma().getSenses().size();j++)
              {
                if(p.getCounts()[i][i]<(2*p.getLemma().getSenses().size()))
                {
                  if(p.getCounts()[i][i]==0)
                  {
                    p.getCounts()[i][j]=1;
                    p.getCounts()[j][i]=1; 
                  }
                  else
                  {
                    p.getCounts()[i][j]=p.getCounts()[i][i];
                    p.getCounts()[j][i]=p.getCounts()[i][i];
                  }                 
                }
                else
                {
                  if(j>i)
                  {
                    p.getCounts()[i][j]=0;
                  }
                }               
              }
            }
         
            for(int i=0;i<p.getLemma().getSenses().size();i++)
            {
              for(int j=i+1;j<p.getLemma().getSenses().size();j++)
              {
                if(p.getCounts()[i][i]>=(2*p.getLemma().getSenses().size())&&p.getCounts()[j][j]>=(2*p.getLemma().getSenses().size()))
                {
                  p.getCounts()[i][j]=(int)wiki.getCounts(p.getLemma().getSenses().get(i),p.getLemma().getSenses().get(j));
                  p.getCounts()[j][i]=p.getCounts()[i][j];
                  if(p.getCounts()[j][i]>=(2*p.getLemma().getSenses().size()))
                  {
                    i=p.getLemma().getSenses().size();
                    break;
 
View Full Code Here
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
   *  This method removes undesired senses from a lemma extracted from Wikipedia.
   *  @param lemma The target lemma.
   */
  public void modifyBow(Lemma lemma) throws Exception {
    ArrayList<Sense> removeList=new ArrayList<Sense>();
    Wiki w=(Wiki)((DataBroker)this.dict).getSource();
    int i=0;
   
    for(Sense s:lemma.getSenses())
    {
      if(w.isNotAnArticle(s.getSid()))
      {
        removeList.add(s);
      }
      i++;
    }
    for(Sense s:removeList)
    {
      lemma.getSenses().remove(s);
    }
    for(Sense s:lemma.getSenses())
    {
      for(ArrayList<Relation> list: s.getRelations().values())
      {
       
        for(int j=0;j<list.size();j++)
        {
          if(w.isNotAnArticle(list.get(j).getSid()))
          {
            list.remove(j);
            j--;
          }
       
View Full Code Here
8
9
10
11
12
13
14
15
16
17
18
19
20
  /**
   * @param args
   */
  public static void main(String[] args)throws Exception{
    Wiki w=new Wiki();
    w.setVersion(args[0]);
    w.setPath(args[1]);
    w.load("all");
    File d=new File(args[3]);
    d.mkdirs();
    w.createInputFromArticle(args[2], args[3]);
  }
View Full Code Here
TOP

Related Classes of gannuNLP.dictionaries.Wiki

  • gannuNLP.corpus.WikiCorpus
  • gannuNLP.data.Count
  • gannuNLP.data.Lemma
  • gannuNLP.data.Relation
  • gannuNLP.data.Sense
  • gannuWSD.bowmodifiers.WikiCleaner
  • gannuWSD.Wiki2SGF
  • org.jdom2.Content
  • org.jdom2.Document
  • org.jdom2.Element
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.