Package uk.ac.cam.ch.wwmm.oscar3.recogniser.document

Examples of uk.ac.cam.ch.wwmm.oscar3.recogniser.document.ProcessingDocument

109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
      }
      File f = new File(fn);
      f = f.getParentFile();
      Document safDoc = new Builder().build(new File(f, "saf.xml"));
      Document sourceDoc = new Builder().build(new File(f, "source.xml"));     
      ProcessingDocument procDoc = ProcessingDocumentFactory.getInstance().makeTokenisedDocument(sourceDoc, false, false, false, safDoc);
      //NameRecogniser nr = new NameRecogniser();
      //nr.halfProcess(sourceDoc);
      //nr.buildTokenTables(safDoc.getRootElement(), false, false);

      for(Entry e : entriesByFileName.get(fn)) {
        Token t = procDoc.getTokenByStart(e.start);
        if(t != null) {
          List<String> features = featuresForToken(t);
          Event event = new Event(e.type, features.toArray(new String[0]));
          events.add(event);
        }
View Full Code Here
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
   
    //NameRecogniser nr = new NameRecogniser();
    //nr.halfProcess(sourceDoc);
    //nr.makeTokenisers(false);

    ProcessingDocument procDoc = ProcessingDocumentFactory.getInstance().makeTokenisedDocument(sourceDoc, false, false, false, safDoc);
   
    Nodes n = safDoc.query("/saf/annot[slot[@name='type']['PRW']]");

    Map<TokenSequence,Boolean> tokSeqs = new HashMap<TokenSequence,Boolean>();
    Map<TokenSequence,Bag<String>> tokSeqPRWs = new HashMap<TokenSequence,Bag<String>>();
   
    for(int i=0;i<n.size();i++) {
      Element e = (Element)n.get(i);
      Token token = procDoc.getTokenByStart(e.getAttributeValue("from"));
      if(token == null) token = procDoc.getTokenByEnd(e.getAttributeValue("to"));
      if(token == null) {
        System.out.println("Eeep!");
      } else {
        TokenSequence tokSeq = token.getTokenSequence();
        //boolean isReact = "REACT".equals(SafTools.getSlotValue(e, "subtype"));
View Full Code Here
29
30
31
32
33
34
35
36
37
38
39
    ResourceGetter rg = new ResourceGetter("uk/ac/cam/ch/wwmm/oscar3/test/testcard/resources/");
    String s = rg.getString("testcard.txt");
    assertTrue("Have testcard string", s != null && s.length() > 0);
    Document doc = TextToSciXML.textToSciXML(s);

    ProcessingDocument procDoc = ProcessingDocumentFactory.getInstance().makeTokenisedDocument(doc, false, false, false);
    assertTrue(procDoc != null);
    List<NamedEntity> neList;
    ChemicalEntityRecogniser cei = new PatternRecogniser();
    neList = cei.findNamedEntities(procDoc);
    assertTrue(neList != null);
View Full Code Here
29
30
31
32
33
34
35
36
37
38
39
    ResourceGetter rg = new ResourceGetter("uk/ac/cam/ch/wwmm/oscar3/test/testcard/resources/");
    String s = rg.getString("testcard.txt");
    assertTrue("Have testcard string", s != null && s.length() > 0);
    Document doc = TextToSciXML.textToSciXML(s);
   
    ProcessingDocument procDoc = ProcessingDocumentFactory.getInstance().makeTokenisedDocument(doc, false, false, false);
    assertTrue(procDoc != null);
    List<NamedEntity> neList;
    ChemicalEntityRecogniser cei = new MEMMRecogniser();
    neList = cei.findNamedEntities(procDoc);
    assertTrue(neList != null);
View Full Code Here
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
            Element e = (Element)n.get(i);
            e.addAttribute(new Attribute("type", "CHEMICAL"));
            //XOMTools.removeElementPreservingText((Element)n.get(i));
          }         
        }
        ProcessingDocument procDoc = ProcessingDocumentFactory.getInstance().makeTokenisedDocument(doc, true, false, false);

        //NameRecogniser nr = new NameRecogniser();
        //nr.halfProcess(doc);

        //if(memm.patternFeatures) {
        //  nr.findForReps(cheatTokenisation);         
        //} else {
          //nr.makeTokenisers(cheatTokenisation);
        //}
       
        int paperTestNEs = 0;
       
        List<Double> paperGoodProbs = new ArrayList<Double>();
        List<Double> paperBadProbs = new ArrayList<Double>();

        Map<NamedEntity,Double> confidences = new HashMap<NamedEntity,Double>();
       
        List<NamedEntity> entities;
       
        Set<String> testNEs = new LinkedHashSet<String>();

        for(TokenSequence tokSeq : procDoc.getTokenSequences()) {
          Nodes neNodes = tokSeq.getElem().query(".//ne");
          for(int k=0;k<neNodes.size();k++) {
            Element neElem = (Element)neNodes.get(k);
            if(filterType != null && !neElem.getAttributeValue("type").equals(filterType)) continue;
            if(antiFilterType != null && neElem.getAttributeValue("type").equals(antiFilterType)) continue;
View Full Code Here
242
243
244
245
246
247
248
249
250
251
252
253
254
255
      }*/
      //NameRecogniser nr = new NameRecogniser();
      //nr.halfProcess(xmlDoc);
      //nr.makeTokenisers(false);
     
      ProcessingDocument procDoc = ProcessingDocumentFactory.getInstance().makeTokenisedDocument(xmlDoc, false, false, false);
     
      //for(Tokeniser t : nr.getTokenisers()) {
        TokenStream ts = new Oscar3TokenStream(procDoc.getTokenSequences());
        ts = new Oscar3Filters(ts);
        luceneDoc.add(new Field("txt", ts, Field.TermVector.WITH_POSITIONS_OFFSETS));
       
      //}
     
View Full Code Here
51
52
53
54
55
56
57
58
59
60
61
   * @param sourceDoc The SciXML document to parse, now unlikely to be altered.
   * @param safDoc An (empty) SAF XML document.
   * @throws Exception C
   */
  private void processDocumentInternal(Document sourceDoc, Document safDoc) throws Exception {
    ProcessingDocument procDoc = ProcessingDocumentFactory.getInstance().makeTokenisedDocument(new Document((Element)XOMTools.safeCopy(sourceDoc.getRootElement())), false, false, false);
    List<NamedEntity> neList;
    ChemicalEntityRecogniser cer = null;
    try
    {
      cer = (ChemicalEntityRecogniser) Class.forName(Oscar3Props.getInstance().chemicalEntityRecogniser).newInstance();
View Full Code Here
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
      //files = FileTools.getFilesFromDirectoryByName(new File("/home/ptc24/oscarworkspace/corpora/paperset1"), "source.xml");
      //files = FileTools.getFilesFromDirectoryByName(new File("/home/ptc24/newows/corpora/paperset1"), "source.xml");
      //files = FileTools.getFilesFromDirectoryByName(new File("/home/ptc24/newows/corpora/BioIE"), "source.xml");
      files = FileTools.getFilesFromDirectoryByName(new File("/home/ptc24/newows/corpora/roughPubMed"), "source.xml");
      //files = FileTools.getFilesFromDirectoryByName(new File("/scratch/pubmed/2005"), "source.xml");
      StringSource ss = new StringSource(files, false);
     
      Bag<String> wordCounts = new Bag<String>();
     
      ss.reset();
      for(String s : ss) {
        TokenSequence t = Tokeniser.getInstance().tokenise(s);
        for(String word : t.getTokenStringList()) {
          if(!word.matches(".*[a-z][a-z].*")) continue;
          word = StringTools.normaliseName(word);
View Full Code Here
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
    Set<String> engWords = new HashSet<String>(NGramBuilder.getInstance().engSet);
   
    List<File> files = new ArrayList<File>();
    files = FileTools.getFilesFromDirectoryByName(new File("/home/ptc24/newows/corpora/roughPubMed"), "source.xml");

    StringSource ss = new StringSource(files, false);
   
    Bag<String> wordCounts = new Bag<String>();
   
    ss.reset();
    for(String s : ss) {
      TokenSequence t = Tokeniser.getInstance().tokenise(s);
      for(String word : t.getTokenStringList()) {
        if(!word.matches(".*[a-z][a-z].*")) continue;
        word = StringTools.normaliseName(word);
View Full Code Here
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
        neByLastToken.put(ne.getLastToken(), ne);
      }
    }

    for(NamedEntity ne : entities) {
      Token prev = ne.getFirstToken().getNAfter(-1);
      Token next = ne.getLastToken().getNAfter(1);
      if(prev != null && next != null && prev.getValue().equals("(") && next.getValue().equals(")")) {
        Token prev2 = ne.getFirstToken().getNAfter(-2);
        if(prev2 != null) {
          String surf = ne.getSurface();
          if(surf.matches(".*[A-Z]s") || prev2.getValue().endsWith("s")) surf = surf.substring(0, surf.length()-1);
          List<String> featuresForAbbrev;
          if(abbrevFeatures.containsKey(surf)) {
            featuresForAbbrev = abbrevFeatures.get(surf);
          } else {
            featuresForAbbrev = new ArrayList<String>();
View Full Code Here
TOP

Related Classes of uk.ac.cam.ch.wwmm.oscar3.recogniser.document.ProcessingDocument

  • dk.brics.automaton.Automaton
  • dk.brics.automaton.RunAutomaton
  • jregex.Pattern
  • nu.xom.Attribute
  • nu.xom.Builder
  • nu.xom.Document
  • nu.xom.Element
  • nu.xom.Elements
  • nu.xom.Node
  • nu.xom.Nodes
    • 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.