-
Notifications
You must be signed in to change notification settings - Fork 0
/
TrainAndTest.java
346 lines (289 loc) · 8.29 KB
/
TrainAndTest.java
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
122
123
124
125
126
127
128
129
130
131
132
133
134
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
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
230
231
232
233
234
235
236
237
238
239
240
241
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
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
import java.util.*;
import java.lang.Math;
import java.io.*;
import java.lang.String;
/**
*
*/
public class TrainAndTest
{
/*
* Parses through review to store term frequencies for each token and increments
* noDocumentsContainToken HashMap if token is found.
*
*/
public HashMap<String,Double> localTokenCount;
public HashMap<String,Double> noDocumentsContainToken;
public HashMap<String,Integer> featureIDMap;
public HashMap<Integer,String> reverseFeatureIDMap;
public double totalNoDocuments;
public int indexCount;
public double maxFrequency;
public TrainAndTest()
{
this.localTokenCount = new HashMap<String,Double>();
this.noDocumentsContainToken = new HashMap<String,Double>();
this.featureIDMap = new HashMap<String,Integer>();
this.reverseFeatureIDMap = new HashMap<Integer,String>();
this.totalNoDocuments = 0;
this.indexCount=1;
}
public void tfIDf( ArrayList<String> tokens )
{
double tokenFrequency, documentsContainingToken;
double inverseDocumentFrequency;
Collections.sort( tokens );
String prevToken="", token;
for( int k=0; k<tokens.size() ;k++ )
{
token = tokens.get( k );
assert token!=null;
assert noDocumentsContainToken != null;
documentsContainingToken = this.noDocumentsContainToken.get( token );
tokenFrequency = this.localTokenCount.get( token );
inverseDocumentFrequency = Math.log( ( this.totalNoDocuments /( 1 + documentsContainingToken ) ) );
if( !prevToken.equals( token ) )
{
this.localTokenCount.put( token, tokenFrequency * inverseDocumentFrequency );
}
prevToken = token;
}
}
public void normalizeTokens( ArrayList<String> tokens )
{
Collections.sort( tokens );
String prevToken="", token;
double tokenFrequencyCount;
for( int k=0; k<tokens.size() ;k++ )
{
token = tokens.get( k );
tokenFrequencyCount = this.localTokenCount.get( token );
if( !prevToken.equals( token ) )
{
this.localTokenCount.put( token, tokenFrequencyCount/this.maxFrequency );
}
prevToken = token;
}
}
public void logScaleTokens( ArrayList<String> tokens )
{
Collections.sort( tokens );
String prevToken="", token;
double tokenFrequencyCount;
for( int k=0; k<tokens.size() ;k++ )
{
token = tokens.get( k );
tokenFrequencyCount = this.localTokenCount.get( token );
if( !prevToken.equals( token ) )
{
this.localTokenCount.put( token, 1 + Math.log( tokenFrequencyCount ) );
}
prevToken = token;
}
}
public void calculateTokenCount( String token )
{
double tokenFrequencyCount;
if( this.localTokenCount.containsKey( token ) )
{
tokenFrequencyCount = this.localTokenCount.get( token );
tokenFrequencyCount++;
}
else
{
tokenFrequencyCount = 1;
}
this.localTokenCount.put( token, tokenFrequencyCount );
// Check for max freq in document
if( tokenFrequencyCount > this.maxFrequency )
{
this.maxFrequency = tokenFrequencyCount ;
}
}
public void calculateTokenPresence( String token )
{
double tokenPresence;
if( !this.localTokenCount.containsKey( token ) )
{
tokenPresence = 1;
this.localTokenCount.put( token, tokenPresence );
}
}
public String createReviewString( ArrayList<Integer> tokenIndices, boolean deceptive )
{
int tokenIndex, prevTokenIndex = 0;
String reviewString,token;
if( deceptive )
{
reviewString = "+1 ";
}
else
{
reviewString = "-1 ";
}
Collections.sort( tokenIndices );
for( int j=0; j<tokenIndices.size(); j++ )
{
tokenIndex = tokenIndices.get( j );
token = this.reverseFeatureIDMap.get( tokenIndex );
if( prevTokenIndex != tokenIndex )
{
reviewString += tokenIndex+":"+this.localTokenCount.get( token )+" ";
}
prevTokenIndex = tokenIndex;
}
reviewString += "\n";
return reviewString;
}
public void buildFeatureVector( String filename, boolean deceptive, String scheme, String tfScheme )
{
String token, reviewLine;
int tokenIndex;
double tokenFrequencyCount;
try
{
FileReader reviewDocumentStream = new FileReader( filename );
BufferedReader reviewDocument = new BufferedReader( reviewDocumentStream );
FileWriter featuresStream = new FileWriter( "features.txt", true );
BufferedWriter featuresOut = new BufferedWriter( featuresStream );
FileWriter inputStream = new FileWriter( "input.txt", true );
BufferedWriter inputOut = new BufferedWriter( inputStream );
while ( ( reviewLine = reviewDocument.readLine() ) != null )
{
Tokenizer tk = new Tokenizer( scheme, reviewLine );
ArrayList<String> tokens = tk.Tokenize() ;
ArrayList<Integer> tokenIndices = new ArrayList<Integer>();
this.maxFrequency=0;
for( int i=0; i<tokens.size() ;i++ )
{
token = tokens.get( i );
if( !this.featureIDMap.containsKey( token ) )
{
tokenIndex = this.indexCount;
this.featureIDMap.put( token, tokenIndex );
this.reverseFeatureIDMap.put( tokenIndex, token );
this.indexCount++;
//Write the feature out to the dictionary
featuresOut.write( token + "\n" );
}
else
{
tokenIndex = this.featureIDMap.get( token );
}
tokenIndices.add( tokenIndex );
if( tfScheme.equals( "RAW_FREQUENCY" )
|| tfScheme.equals( "NORMALIZED_FREQUENCY" )
|| tfScheme.equals( "LOGSCALED_FREQUENCY") )
{
calculateTokenCount( token );
}
else
{
// TOKEN PRESENCE
calculateTokenPresence( token );
}
}
if( tfScheme.equals("NORMALIZED_FREQUENCY") )
{
normalizeTokens( tokens );
}
if( tfScheme.equals("LOGSCALED_FREQUENCY") )
{
logScaleTokens( tokens );
}
tfIDf( tokens );
String reviewString = createReviewString( tokenIndices, deceptive );
inputOut.write( reviewString );
this.localTokenCount.clear();
}
featuresOut.write( "TOTAL NO OF DOCUMENTS : "+this.totalNoDocuments+"" );
inputOut.close();
featuresOut.close();
}
catch ( Exception e )
{
e.printStackTrace();
}
}
public void deletePreviousFiles()
{
try
{
File inputFile = new File( "input.txt" );
File featuresFile = new File( "features.txt" );
if( inputFile.exists() )
{
inputFile.delete();
}
if( featuresFile.exists() )
{
featuresFile.delete();
}
}
catch( Exception e )
{
e.printStackTrace();
}
}
public void DocumentData( ArrayList<String> tokens )
{
String token, prevToken="";
double documentsContainingToken;
Collections.sort( tokens );
for( int i=0; i<tokens.size() ;i++ )
{
token = tokens.get( i );
// Also increment hashmap noDocumentsContainToken for every token encountered
if( this.noDocumentsContainToken.containsKey( token ) )
{
documentsContainingToken = this.noDocumentsContainToken.get( token );
}
else
{
documentsContainingToken = 0;
}
if( !prevToken.equals( token ) )
{
this.noDocumentsContainToken.put( token, documentsContainingToken + 1 );
}
prevToken = token;
}
}
public void documentLevelData( String filename, String scheme )
{
String reviewLine;
try
{
FileReader reviewDocumentStream = new FileReader( filename );
BufferedReader reviewDocument = new BufferedReader( reviewDocumentStream );
while ( ( reviewLine = reviewDocument.readLine() ) != null )
{
Tokenizer tk = new Tokenizer( scheme, reviewLine );
ArrayList<String> tokens = tk.Tokenize() ;
DocumentData( tokens );
this.totalNoDocuments++;
}
}
catch( Exception e )
{
e.printStackTrace();
}
}
public static void main( String[] args )
{
if( args.length < 2 )
{
System.out.println("Insufficient Arguments\n SYNTAX : java TrainAndTest <tokenization_scheme> <tfScheme>");
System.exit(-1);
}
String scheme = args[0];
String tfScheme = args[1]; // RAW_FREQUENCY, NORMALIZED_FREQUENCY, LOGSCALED_FREQUENCY, TOKEN_PRESENCE
TrainAndTest tt = new TrainAndTest();
String truthful = "hotel_truthful_new", deceptive = "hotel_deceptive_new";
tt.documentLevelData( truthful, scheme );
tt.documentLevelData( deceptive, scheme );
tt.deletePreviousFiles();
tt.buildFeatureVector( truthful, false, scheme, tfScheme );
tt.buildFeatureVector( deceptive, true, scheme, tfScheme );
}
}