Package opennlp.tools.parser

Examples of opennlp.tools.parser.Parse

53
54
55
56
57
58
59
60
61
62
63
64
    POSTaggerME tagger =  new POSTaggerME(posModel);
    Parser parser = new ChunkParser(chunker, tagger);
    Parse[] results = ParserTool.parseLine("The Minnesota Twins , " +
            "the 1991 World Series Champions , are currently in third place .",
            parser, 1);
    Parse p = results[0];
    Parse[] chunks = p.getChildren();
    assertTrue(chunks.length == 9);
    assertTrue(chunks[0].getType().equals("NP"));
    assertTrue(chunks[0].getHead().toString().equals("Twins"));
    //<end id="openChunkParse"/>
  }
View Full Code Here
66
67
68
69
70
71
72
73
74
75
76
  @Override
  public Query parse() throws ParseException {

    //<start id="qqp.parse"/>
    Parse parse = ParserTool.parseLine(qstr, parser, 1)[0];//<co id="qqp.parseLine"/>
    /*
    <calloutlist>
        <callout arearefs="qqp.parseLine"><para>Parse the question using the <classname>TreebankParser</classname>.  The resulting <classname>Parse</classname> object can then be utilized by the classifier to determine the Answer Type.</para></callout>
    </calloutlist>
    */
 
View Full Code Here
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
    List<Parse> parts = new ArrayList<Parse>();
    parts.add(parse);
    while (parts.size() > 0) {
      List<Parse> newParts = new ArrayList<Parse>();
      for (int pi=0,pn=parts.size();pi<pn;pi++) {
        Parse cp = parts.get(pi);
        if (cp.getType().equals("NP") && cp.isFlat()) {
          nps.add(cp);
        }
        else if (!cp.isPosTag()) {
          newParts.addAll(Arrays.asList(cp.getChildren()));
        }
      }
      parts = newParts;
    }
    return nps.toArray(new Parse[nps.size()]);
View Full Code Here
112
113
114
115
116
117
118
119
120
121
122
    }
    return nps.toArray(new Parse[nps.size()]);
  }
 
  private Parse getContainingNounPhrase(Parse token) {
    Parse parent = token.getParent();
    if (parent.getType().equals("NP")) {
      return parent;
    }
    return null;
  }
View Full Code Here
121
122
123
124
125
126
127
128
129
130
131
    return null;
  }
 
  private int getTokenIndexFollowingPhrase(Parse p,Parse[] toks) {
    Parse[] ptok = p.getTagNodes();
    Parse lastToken = ptok[ptok.length-1];
    for (int ti=0,tl=toks.length;ti<tl;ti++) {
      if (toks[ti] == lastToken) {
        return(ti+1);
      }
    }
View Full Code Here
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
 
  private Parse findFocusNounPhrase(String queryWord, int qwi, Parse[] toks) {
    if (queryWord.equals("who")) {
      int npStart = movePastCopula(qwi + 1, toks);
      if (npStart > qwi + 1) { // check to ensure there is a copula
        Parse np = getContainingNounPhrase(toks[npStart]);
        if (np != null) {
          return (np);
        }
      }
    }
    else if (queryWord.equals("what")) {
      int npStart = movePastCopula(qwi + 1, toks);
      Parse np = getContainingNounPhrase(toks[npStart]);
      //removed copula case
      if (np != null) {
        Parse head = np.getHead();
        if (falseHeadsPattern.matcher(head.toString()).matches()) {
          npStart += np.getChildren().length;
          int np2Start = movePastPrep(npStart, toks);
          if (np2Start > npStart) {
            Parse snp = getContainingNounPhrase(toks[np2Start]);
            if (snp != null) {
              return (snp);
            }
          }
        }
        return (np);
      }
    }
    else if (queryWord.equals("which")) {
      //check for false query words like which VBD
      int npStart = movePastCopula(qwi + 1, toks);
      if (npStart > qwi + 1) {
        return (getContainingNounPhrase(toks[npStart]));
      }
      else {
        npStart = movePastOf(qwi + 1, toks);
        return (getContainingNounPhrase(toks[npStart]));
      }
    }
    else if (queryWord.equals("how")) {
      if (qwi + 1 < toks.length) {
        return (getContainingNounPhrase(toks[qwi + 1]));
      }
    }
    else if (qwi == 0 && queryWord.equals("name")) {
      int npStart = qwi + 1;
      Parse np = getContainingNounPhrase(toks[npStart]);
      if (np != null) {
        Parse head = np.getHead();
        if (falseHeadsPattern.matcher(head.toString()).matches()) {
          npStart += np.getChildren().length;
          int np2Start = movePastPrep(npStart, toks);
          if (np2Start > npStart) {
            Parse snp = getContainingNounPhrase(toks[np2Start]);
            if (snp != null) {
              return (snp);
            }
          }
        }
View Full Code Here
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
    features.add("hw=" + toks[nsi]);
    features.add("ht=" + toks[nsi].getType());
  }
 
  public String[] getContext(Parse query) {
    Parse focalNoun = null;
    String queryWord = null;
    List<String> features = new ArrayList<String>();
    features.add("def");
    Parse[] nps = getNounPhrases(query);
    Parse[] toks = query.getTagNodes();
    int fnEnd = 0;
    int i = 0;
    boolean fnIsLast = false;
    for (; i < toks.length; i++) {
      String tok = toks[i].toString().toLowerCase();
      if (isQueryWord(tok)) {
        queryWord = tok;
        focalNoun = findFocusNounPhrase(queryWord, i, toks);
        if (tok.equals("how") && i+1 < toks.length) {
          if (howModifierTagPattern.matcher(toks[i+1].getType()).find()) {
            queryWord=tok+"_"+toks[i+1].toString();
          }
        }
        if (focalNoun != null) {
          fnEnd = getTokenIndexFollowingPhrase(focalNoun,toks);
        }
        if (focalNoun != null && focalNoun.equals(nps[nps.length - 1])) {
          fnIsLast  = true;
        }
        break;
      }
    }
View Full Code Here
80
81
82
83
84
85
86
87
 
  public Event next() {
    int split = line.indexOf(' ');
    String outcome = line.substring(0,split);
    String question = line.substring(split+1);
    Parse query = ParserTool.parseLine(question,parser,1)[0];
    return (new Event(outcome, atcg.getContext(query)));
  }
View Full Code Here
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
89
90
91
92
93
      words[i] = children[i].toString();//<co id="cp.words"/>
    }
    String[] tags = tagger.tag(words);//<co id="cp.tag"/>
    tagger.probs(probs);//<co id="cp.probs"/>
    for (int j = 0; j < words.length; j++) {
      Parse word = children[j];
      double prob = probs[j];
      tokens.insert(new Parse(word.getText(), word.getSpan(), tags[j], prob, j));//<co id="cp.augment"/>
      tokens.addProb(Math.log(prob));
    }
    /*
    <calloutlist>
        <callout arearefs="cp.child"><para>The <methodname>parse</methodname> is a callback method from an internal OpenNLP API that tokenizes the original text.</para></callout>
        <callout arearefs="cp.words"><para>Get just the words for use with the tagger</para></callout>
        <callout arearefs="cp.tag"><para>Part of speech tag the words</para></callout>
        <callout arearefs="cp.probs"><para></para></callout>
        <callout arearefs="cp.augment"><para>Augment the initial parse with the part of speech information</para></callout>
    </calloutlist>
    */
    //<end id="cp.pos"/>
    String[] chunks = chunker.chunk(words, tags);
    chunker.probs(probs);
    int chunkStart = -1;
    String chunkType = null;
    double logProb=0;
    for (int ci=0,cn=chunks.length;ci<cn;ci++) {
      if (ci > 0 && !chunks[ci].startsWith("I-") && !chunks[ci-1].equals("O")) {
        Span span = new Span(children[chunkStart].getSpan().getStart(),children[ci-1].getSpan().getEnd());
        tokens.insert(new Parse(tokens.getText(), span, chunkType, logProb,children[ci-1]));
        logProb=0;
      }           
      if (chunks[ci].startsWith("B-")) {
        chunkStart = ci;
        chunkType = chunks[ci].substring(2);
      }
      logProb+=Math.log(probs[ci]);
    }
    if (!chunks[chunks.length-1].equals("O")) {
      int ci = chunks.length;
      Span span = new Span(children[chunkStart].getSpan().getStart(),children[ci-1].getSpan().getEnd());
      tokens.insert(new Parse(tokens.getText(), span, chunkType, logProb,children[ci-1]));
    }
    return tokens;
  }
View Full Code Here
54
55
56
57
58
59
60
61
62
63
64
65
66
67
      tokenSpans[i] = new Span(startOffset, textBuilder.length());
    }
   
    String text = textBuilder.toString();
   
    Parse p = new Parse(text, new Span(0, text.length()), AbstractBottomUpParser.INC_NODE, 0, 0);
   
    for (int i = 0; i < tokenSpans.length; i++) {
      Span tokenSpan = tokenSpans[i];
      p.insert(new Parse(text, new Span(tokenSpan.getStart(), tokenSpan.getEnd()), AbstractBottomUpParser.TOK_NODE, 0, i));
    }
   
    return p;
  }
View Full Code Here
TOP

Related Classes of opennlp.tools.parser.Parse

  • com.tamingtext.opennlp.ChunkParserTest
  • com.tamingtext.qa.AnswerTypeContextGenerator
  • com.tamingtext.qa.AnswerTypeEventStream
  • com.tamingtext.qa.ChunkParser
  • com.tamingtext.qa.QuestionQParser
  • functionality.Parsing
  • net.sf.nlpshell.CorefParse
  • net.sf.nlpshell.Main
  • opennlp.tools.cmdline.coref.CoreferencerTool
  • opennlp.tools.cmdline.coref.CoreferencerTool$CorefParse
    • 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.