1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
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
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
|
package com.wcohen.ss;
import java.io.IOException;
import com.wcohen.ss.api.StringWrapper;
import com.wcohen.ss.abbvGapsHmm.*;
/**
* Abbreviation distance metric which evaluates the probability of a short-form string being an abbreviation/acronym
* of another long-form string. The probability is given by an HMM-based alignment between the two strings.
* <br><br>
* Sample command line:<br>
* <code> java com.wcohen.ss.AbbreviationAlignment "DNA" "Deoxyribonucleic acid" </code><br>
* Expected output:<br>
* <pre>
* M |M |M|M |
* D |N | |A |
* Deoxyribo|nucleic| |acid|
*
* Probability = 3.674595157620664E-6
* </pre>
* where, the rows of the alignment contain: (1) the states of the Viterbi path in the Alignment-HMM, (2) the short form characters, and (3)
* the long form characters. The probability of the Viterbi path is also given.
* <br><br>
* Citation: Dana Movshovitz-Attias and William Cohen, Alignment-HMM-based Extraction of Abbreviations from Biomedical Text, 2012, BioNLP in NAACL
*
* @see com.wcohen.ss.expt.ExtractAbbreviations
* @author Dana Movshovitz-Attias
*
*/
public class AbbreviationAlignment extends AbstractStatisticalTokenDistance {
private AlignmentPredictionModel _alignPredictor = null;
private static String _trainingDir = "./train/";
/**
* Evaluates the probability of the short-form string (string1) being an abbreviation/acronym
* of the long-form string (string2).<br>
* Usage: AbbreviationAlignment short_form_string long_form_string [train_data_dir]
*/
static public void main(String[] argv) {
String[] newArgv = init(argv);
AbbreviationAlignment aligner = new AbbreviationAlignment();
doMain(aligner, newArgv);
}
static public String[] init(String[] argv) {
if(argv.length < 2){
System.out.println("Usage: AlignmentPredictionModel short_form_string long_form_string [train_data_dir]\n\n"+
"short_form_string long_form_string - Candidate abbreviation strings, for example, \"DNA\" \"Deoxyribonucleic acid\"\n"+
"train_data_dir - Optional. Directory containing a corpus file named 'abbvAlign_corpus.txt' for training the abbreviation HMM. "+
"Corpus format is one line per file.\n"+
" The model parameters will be saved in this directory under 'hmmModelParams.txt' so the HMM will only have to be trained once.\n"+
" Default = './train/'\n\n"+
"Example: java com.wcohen.ss.AbbreviationAlignment \"DNA\" \"Deoxyribonucleic acid\"\n"+
"Expected output:\n"+
"M |M |M|M |\n"+
"D |N | |A |\n"+
"Deoxyribo|nucleic| |acid|\n"+
"\n"+
"Probability = 3.674595157620664E-6");
System.exit(1);
}
if(argv.length >= 3) {
AbbreviationAlignment._trainingDir = argv[2];
}
String[] newArgv = {argv[0], argv[1]};
return newArgv;
}
public AbbreviationAlignment() {
loadPredictor();
setTrainDir(_trainingDir);
}
public void setTrainDir(String trainDir) {
_trainingDir = trainDir;
_alignPredictor.setTrainingDataDir(trainDir+"/");
_alignPredictor.setModelParamsFile(trainDir+"/hmmModelParams.txt");
_alignPredictor.trainIfNeeded();
}
protected AlignmentPredictionModel loadPredictor(){
if(_alignPredictor == null){
try {
_alignPredictor = new AlignmentPredictionModel();
} catch (IOException e) {
System.err.println("Unable to load AlignmentPredictionModel");
e.printStackTrace();
System.exit(1);
}
}
return _alignPredictor;
}
@Override
public double score(StringWrapper s, StringWrapper t) {
com.wcohen.ss.abbvGapsHmm.AbbreviationAlignmentContainer<AbbvGapsHMM.Emissions, AbbvGapsHMM.States> alignment = _alignPredictor.predict(s.unwrap(), t.unwrap());
if(alignment != null){
return alignment.getProbability();
}
return 0;
}
@Override
public String explainScore(StringWrapper s, StringWrapper t) {
String explainStr = "";
com.wcohen.ss.abbvGapsHmm.AbbreviationAlignmentContainer<AbbvGapsHMM.Emissions, AbbvGapsHMM.States> alignment = _alignPredictor.predict(s.unwrap(), t.unwrap());
if(alignment != null){
explainStr += alignment.toString()+"\n\n";
explainStr += "Probability = "+alignment.getProbability();
}
else{
explainStr += "No alignment found between: \""+s.unwrap()+"\", \""+t.unwrap()+"\"";
}
return explainStr;
}
}
|
