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

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

88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
      }
      prevCell = cell;
    }
    //System.out.println("**************************");
    // Parenthetical phrases
    Token beforeBracket = null;
    Token prevToken = null;
    for(Token token : tokSeq.getTokens()) {
      if(prevToken == null) {
        prevToken = token;
        continue;
      }
      if(beforeBracket == null) {
        if(token.getValue().equals("(")) beforeBracket = prevToken;
      } else if(token.getValue().equals(")")) {
        LatticeCell beforeBracketCell = tokensToTokenCells.get(beforeBracket);
        LatticeCell bracketCell = tokensToTokenCells.get(token);
        if(beforeBracketCell != null || bracketCell != null) {
          beforeBracketCell.addInheritance(bracketCell);
          //System.out.println("Parenthetical: " + tokSeq.getSubstring(beforeBracket.getId(), token.getId()));
        }
        beforeBracket = null;
      }
      prevToken = token;
    }
   
    // Build in named entities
    for(Token token : tokSeq.getTokens()) {
      if(startsToNEs.containsKey(token.getStart())) {
        for(Element neElem : startsToNEs.get(token.getStart())) {
          Token endToken = procDoc.getTokenByEnd(neElem.getAttributeValue("to"));
          if(endToken == null) continue;
          prevCell = tokenToPrev.get(token);
          LatticeCell endTokenCell = tokensToTokenCells.get(endToken);
          if(endTokenCell == null) continue;
          LatticeCell neCell = new LatticeCell(neElem, endTokenCell);
View Full Code Here
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
 
  public static List<String> featuresForToken(Token token) {
    List<String> features = new ArrayList<String>();
    for(int i=1;i<=10;i++) {
      int inum = i;
      Token after = token.getNAfter(i);
      if(after != null) features.add("afters" + inum + "=" + st.getStem(after.getValue()));
      if(after != null) features.add("afters" + (inum+1) + "=" + st.getStem(after.getValue()));
      //if(after != null) features.add("after" + inum + "=" + after.getValue());
      //if(after != null) features.add("after" + (inum+1) + "=" + after.getValue());
      if(after != null) features.add("ui=" + after.getValue());
      Token before = token.getNAfter(-i);
      if(before != null) features.add("befores" + inum + "=" + st.getStem(before.getValue()));
      if(before != null) features.add("befores" + (inum+1) + "=" + st.getStem(before.getValue()));
      //if(before != null) features.add("before" + inum + "=" + before.getValue());
      //if(before != null) features.add("before" + (inum+1) + "=" + before.getValue());
      if(before != null) features.add("ui=" + before.getValue());
    }
   
    return features;
  }
View Full Code Here
115
116
117
118
119
120
121
122
123
124
125
      //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
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
    for(NamedEntity ne : neList) {
      if(ne.getType().equals(NETypes.POTENTIALACRONYM)) {
        int start = ne.getStart();
        //int end = ne.getEnd();

        Token t = tokensByStart.get(start);
        //if(t != null) System.out.println("AHA: " + t.getValue());
        if(t != null && t.getNAfter(-2) != null && t.getNAfter(1) != null) {
          Token prev = t.getNAfter(-1);
          Token next = t.getNAfter(1);
          Token prevPrev = t.getNAfter(-2);
          if(prev.getValue().equals("(") && next.getValue().endsWith(")")) {
            //boolean matched = false;
            if(endToNe.containsKey(prevPrev.getEnd())) {
              NamedEntity acronymOf = endToNe.get(prevPrev.getEnd());
              if(StringTools.testForAcronym(ne.getSurface(), acronymOf.getSurface())) {
                //System.out.println(ne.getSurface() + " is " + acronymOf.getSurface());
                if(acronymOf.getType().equals(NETypes.ASE) || acronymOf.getType().equals(NETypes.ASES)) {
                  //System.out.println("Skip ASE acronym");
                } else {
View Full Code Here
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
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
    boolean doSort = true;
    for(String s : ss) {
      s = s.replaceAll("\\s+", " ");
      TokenSequence t = Tokeniser.getInstance().tokenise(s);
      for(int i=0;i<t.size();i++) {
        Token token = t.getToken(i);
        if(token.getValue().equalsIgnoreCase(word1)) {
          for(int j=i+1;j<t.size() && j<i+10;j++) {
            Token token2 = t.getToken(j);
            if(token2.getValue().equalsIgnoreCase(word2)) {
              int wstart1 = token.getStart();
              int wend1 = token.getEnd();
             
              int wstart2 = token2.getStart();
              int wend2 = token2.getEnd();
             
              String before = null;
              if(wstart1 < beforeWidth) {
                before = StringTools.multiplyString(" ", beforeWidth - wstart1) + s.substring(0, wstart1);
              } else {
                before = s.substring(wstart1 - beforeWidth, wstart1);
              }
             
              String after = null;
              if(s.length() - wend2 > afterWidth) {
                after = s.substring(wend2, wend2 + afterWidth);
              } else {
                after = s.substring(wend2);
              }
             
              String middle = s.substring(wend1, wstart2);
              String middleDisplay = middle;
              if(middle.length() > middleWidth) continue;
              if(middle.length() < middleWidth) {
                if(mode.startsWith("left")) {
                  middleDisplay += StringTools.multiplyString(" ", middleWidth - middle.length());                                   
                } else if(mode.startsWith("right")) {
                  middleDisplay = StringTools.multiplyString(" ", middleWidth - middle.length()) + middleDisplay;                 
                } else {
                  int toPad = middleWidth - middle.length();
                  int left = toPad / 2;
                  int right = toPad - left;
                  middleDisplay = StringTools.multiplyString(" ", left) +
                    middleDisplay + StringTools.multiplyString(" ", right);
                }
              }
             
              String display = before + "  " + token.getValue() + "  " + middleDisplay + " " + token2.getValue() + " " + after;
              String sort = "";
              if("interleave".equals(mode)) {
                sort = middleDisplay.toLowerCase() + interleave(before, after, true);           
              } else if("left".equals(mode)) {
                StringBuffer sb = new StringBuffer(before.toLowerCase());
View Full Code Here
104
105
106
107
108
109
110
111
   *
   * @param word The string to test.
   */
  public static void destroyInstanceIfWordTokenises(String word) {
    if(myInstance == null) return;
    TokenSequence ts = Tokeniser.getInstance().tokenise(word);
    if(ts.getTokens().size() > 1) myInstance = null;
  }
View Full Code Here
127
128
129
130
131
132
133
134
   *
   * @param word The string to test.
   */
  public static void destroyInstanceIfWordTokenises(String word) {
    if(myInstance == null) return;
    TokenSequence ts = Tokeniser.getInstance().tokenise(word);
    if(ts.getTokens().size() > 1) myInstance = null;
  }
View Full Code Here
133
134
135
136
137
138
139
140
141
142
143
144
        ontDFANumber++;
      }
      type = type + "_" + ontDFANumber;*/
    }

    TokenSequence ts = Tokeniser.getInstance().tokenise(ne);
    List<String> tokens = ts.getTokenStringList();

    if(!alwaysAdd && tokens.size() == 1 && !ne.contains("$")) return;
    StringBuffer sb = new StringBuffer();
    for(String token : tokens) {
      sb.append(getRepForToken(token));
View Full Code Here
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
                featuresForAbbrev.add(abbrMode + "abbr<" + thresh);
              }             
            }
          } else {
            int tokID = ne.getFirstToken().getId();
            TokenSequence tokSeq = ne.getFirstToken().getTokenSequence();
            int length = surf.length();
            boolean isAcro = false;
            if(allCaps.matcher(surf).matches()) {
              if(length <= (tokID - 1)) {
                isAcro = true;
                for(int i=0;i<length;i++) {
                  if(!tokSeq.getToken(tokID - length - 1 + i).getValue().toUpperCase().startsWith(surf.substring(i,i+1))) isAcro = false;
                }
                if(isAcro) {
                  featuresForAbbrev.add("allUpperAbbrev");
                }
              }
View Full Code Here
126
127
128
129
130
131
132
133
134
135
136
    double conf = ne.getConfidence();
    double confLog = Math.log(conf) - Math.log(1 - conf);

    List<String> features = new ArrayList<String>();
   
    TokenSequence t = ne.getTokens().get(0).getTokenSequence();
    int entityLength = ne.getTokens().size();
    int startID = ne.getTokens().get(0).getId();
    int endID = startID + entityLength - 1;
    String surf = ne.getSurface();
   
View Full Code Here
TOP

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

  • 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.