line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
1
|
|
|
|
|
|
|
#!/usr/bin/perl |
2
|
|
|
|
|
|
|
# NanoB2B-NER::NER::Avgman |
3
|
|
|
|
|
|
|
# |
4
|
|
|
|
|
|
|
# Averages Weka files and makes a nice output file |
5
|
|
|
|
|
|
|
# Version 1.5 |
6
|
|
|
|
|
|
|
# |
7
|
|
|
|
|
|
|
# Program by Milk |
8
|
|
|
|
|
|
|
|
9
|
|
|
|
|
|
|
package NanoB2B::NER::Avgman; |
10
|
|
|
|
|
|
|
|
11
|
1
|
|
|
1
|
|
6
|
use NanoB2B::UniversalRoutines; |
|
1
|
|
|
|
|
2
|
|
|
1
|
|
|
|
|
27
|
|
12
|
1
|
|
|
1
|
|
4
|
use File::Path qw(make_path); #makes sub directories |
|
1
|
|
|
|
|
2
|
|
|
1
|
|
|
|
|
35
|
|
13
|
1
|
|
|
1
|
|
5
|
use strict; |
|
1
|
|
|
|
|
1
|
|
|
1
|
|
|
|
|
16
|
|
14
|
1
|
|
|
1
|
|
4
|
use warnings; |
|
1
|
|
|
|
|
2
|
|
|
1
|
|
|
|
|
2227
|
|
15
|
|
|
|
|
|
|
|
16
|
|
|
|
|
|
|
#### GLOBAL VARIABLES #### |
17
|
|
|
|
|
|
|
|
18
|
|
|
|
|
|
|
#option variables |
19
|
|
|
|
|
|
|
my $program_dir; |
20
|
|
|
|
|
|
|
my $bucketsNum = 10; |
21
|
|
|
|
|
|
|
my $weka_dir; |
22
|
|
|
|
|
|
|
my $debug = 0; |
23
|
|
|
|
|
|
|
|
24
|
|
|
|
|
|
|
#universal subroutines object |
25
|
|
|
|
|
|
|
my %uniParams; |
26
|
|
|
|
|
|
|
my $uniSub; |
27
|
|
|
|
|
|
|
|
28
|
|
|
|
|
|
|
#module variables |
29
|
|
|
|
|
|
|
my @features; |
30
|
|
|
|
|
|
|
my @wekaAttr = ("TP Rate", "FP Rate", "Precision", "Recall", "F-Measure", "MCC", "ROC Area", "PRC Area"); |
31
|
|
|
|
|
|
|
my @allBuckets; |
32
|
|
|
|
|
|
|
my @sets; |
33
|
|
|
|
|
|
|
my $entId = "_e"; |
34
|
|
|
|
|
|
|
|
35
|
|
|
|
|
|
|
#### A SUPER PET IS ADOPTED #### |
36
|
|
|
|
|
|
|
|
37
|
|
|
|
|
|
|
# construction method to create a new Avgman object |
38
|
|
|
|
|
|
|
# input : $directory <-- the name of the directory for the files |
39
|
|
|
|
|
|
|
# $weka_dir <-- the weka directory name |
40
|
|
|
|
|
|
|
# $features <-- the set of features [e.g. "ortho morph text pos cui sem"] |
41
|
|
|
|
|
|
|
# $buckets <-- the number of buckets used for the k-fold cross validation |
42
|
|
|
|
|
|
|
# $debug <-- the set of features to run on [e.g. omtpcs] |
43
|
|
|
|
|
|
|
# output : $self <-- an instance of the Avgman object |
44
|
|
|
|
|
|
|
sub new { |
45
|
|
|
|
|
|
|
#grab class and parameters |
46
|
0
|
|
|
0
|
0
|
|
my $self = {}; |
47
|
0
|
|
|
|
|
|
my $class = shift; |
48
|
0
|
0
|
|
|
|
|
return undef if(ref $class); |
49
|
0
|
|
|
|
|
|
my $params = shift; |
50
|
|
|
|
|
|
|
|
51
|
|
|
|
|
|
|
#reset all the arrays |
52
|
0
|
|
|
|
|
|
%uniParams = (); |
53
|
0
|
|
|
|
|
|
@features = (); |
54
|
0
|
|
|
|
|
|
@allBuckets = (); |
55
|
0
|
|
|
|
|
|
@sets = (); |
56
|
|
|
|
|
|
|
|
57
|
|
|
|
|
|
|
#bless this object |
58
|
0
|
|
|
|
|
|
bless $self, $class; |
59
|
0
|
|
|
|
|
|
$self->_init($params); |
60
|
0
|
|
|
|
|
|
@allBuckets = (1..$bucketsNum); |
61
|
|
|
|
|
|
|
|
62
|
|
|
|
|
|
|
#retrieve parameters for universal-routines |
63
|
0
|
|
|
|
|
|
$uniParams{'debug'} = $debug; |
64
|
0
|
|
|
|
|
|
$uniSub = NanoB2B::UniversalRoutines->new(\%uniParams); |
65
|
|
|
|
|
|
|
|
66
|
|
|
|
|
|
|
#make the features |
67
|
0
|
|
|
|
|
|
my $item = "_"; |
68
|
0
|
|
|
|
|
|
foreach my $fs (@features){ |
69
|
0
|
|
|
|
|
|
$item .= substr($fs, 0, 1); #add to abbreviations for the name |
70
|
0
|
|
|
|
|
|
push(@sets, $item); |
71
|
|
|
|
|
|
|
} |
72
|
|
|
|
|
|
|
|
73
|
|
|
|
|
|
|
|
74
|
0
|
|
|
|
|
|
return $self; |
75
|
|
|
|
|
|
|
} |
76
|
|
|
|
|
|
|
|
77
|
|
|
|
|
|
|
# method to initialize the NanoB2B::NER::Avgman object. |
78
|
|
|
|
|
|
|
# input : $parameters <- reference to a hash |
79
|
|
|
|
|
|
|
# output: |
80
|
|
|
|
|
|
|
sub _init { |
81
|
0
|
|
|
0
|
|
|
my $self = shift; |
82
|
0
|
|
|
|
|
|
my $params = shift; |
83
|
|
|
|
|
|
|
|
84
|
0
|
0
|
|
|
|
|
$params = {} if(!defined $params); |
85
|
|
|
|
|
|
|
|
86
|
|
|
|
|
|
|
# get some of the parameters |
87
|
0
|
|
|
|
|
|
my $diroption = $params->{'directory'}; |
88
|
0
|
|
|
|
|
|
my $ftsoption = $params->{'features'}; |
89
|
0
|
|
|
|
|
|
my $bucketsNumoption = $params->{'buckets'}; |
90
|
0
|
|
|
|
|
|
my $wekadiroption = $params->{'weka_dir'}; |
91
|
0
|
|
|
|
|
|
my $debugoption = $params->{'debug'}; |
92
|
|
|
|
|
|
|
|
93
|
|
|
|
|
|
|
#set the global variables |
94
|
0
|
0
|
|
|
|
|
if(defined $debugoption){$debug = $debugoption;} |
|
0
|
|
|
|
|
|
|
95
|
0
|
0
|
|
|
|
|
if(defined $diroption){$program_dir = $diroption;} |
|
0
|
|
|
|
|
|
|
96
|
0
|
0
|
|
|
|
|
if(defined $bucketsNumoption){$bucketsNum = $bucketsNumoption;} |
|
0
|
|
|
|
|
|
|
97
|
0
|
0
|
|
|
|
|
if(defined $ftsoption){@features = split(' ', $ftsoption);} |
|
0
|
|
|
|
|
|
|
98
|
0
|
0
|
|
|
|
|
if(defined $wekadiroption){$weka_dir = $wekadiroption}; |
|
0
|
|
|
|
|
|
|
99
|
|
|
|
|
|
|
} |
100
|
|
|
|
|
|
|
|
101
|
|
|
|
|
|
|
|
102
|
|
|
|
|
|
|
############### NOT YOUR AVGMAN! ################ |
103
|
|
|
|
|
|
|
|
104
|
|
|
|
|
|
|
#prints weka results to an output file |
105
|
|
|
|
|
|
|
# input : $name <-- name of the file with the weka data |
106
|
|
|
|
|
|
|
# output : (weka average results file) |
107
|
|
|
|
|
|
|
sub avg_file{ |
108
|
0
|
|
|
0
|
0
|
|
my $self = shift; |
109
|
0
|
|
|
|
|
|
my $name = shift; |
110
|
|
|
|
|
|
|
|
111
|
|
|
|
|
|
|
#get the averaged weka data, individual weka data, and majority sense baseline |
112
|
0
|
|
|
|
|
|
my %avg = averageWekaData($name); |
113
|
0
|
|
|
|
|
|
my %ind = individualWekaData($name); |
114
|
0
|
|
|
|
|
|
my $msb = getMSB($name); |
115
|
|
|
|
|
|
|
|
116
|
|
|
|
|
|
|
#open new file |
117
|
0
|
|
|
|
|
|
my $direct = ("$program_dir/_WEKAS/$weka_dir"); |
118
|
0
|
|
|
|
|
|
make_path($direct); |
119
|
0
|
|
|
|
|
|
my $file = "$direct/$name" . "_weka-results"; |
120
|
0
|
|
|
|
|
|
my $WEKA_SAVE; |
121
|
0
|
0
|
|
|
|
|
open ($WEKA_SAVE, ">", "$file") || die ("Weka file $file could not be made in the _WEKAS directory! \n$!"); |
122
|
|
|
|
|
|
|
|
123
|
|
|
|
|
|
|
#print the name |
124
|
0
|
|
|
|
|
|
$uniSub->print2File($WEKA_SAVE, "ENTITY NAME\t\t: $name"); |
125
|
|
|
|
|
|
|
|
126
|
|
|
|
|
|
|
#print the list of features |
127
|
0
|
|
|
|
|
|
$uniSub->print2FileNoLine($WEKA_SAVE, "FEATURES\t\t: "); |
128
|
0
|
|
|
|
|
|
foreach my $d(@features){$uniSub->print2FileNoLine($WEKA_SAVE, "$d, ");} |
|
0
|
|
|
|
|
|
|
129
|
0
|
|
|
|
|
|
$uniSub->print2File($WEKA_SAVE, ""); |
130
|
|
|
|
|
|
|
|
131
|
|
|
|
|
|
|
#print the msb |
132
|
0
|
|
|
|
|
|
$msb = int($msb * 10**4) / 10**4; |
133
|
0
|
|
|
|
|
|
$uniSub->print2File($WEKA_SAVE, "Majority Label Baseline : $msb"); |
134
|
|
|
|
|
|
|
|
135
|
0
|
|
|
|
|
|
$uniSub->print2File($WEKA_SAVE, "\n"); |
136
|
|
|
|
|
|
|
|
137
|
|
|
|
|
|
|
#print the averages |
138
|
0
|
|
|
|
|
|
foreach my $key (sort keys %avg){ |
139
|
0
|
|
|
|
|
|
$uniSub->print2File($WEKA_SAVE, "$key - AVERAGES"); |
140
|
0
|
|
|
|
|
|
$uniSub->print2File($WEKA_SAVE, "-------------------------"); |
141
|
|
|
|
|
|
|
|
142
|
0
|
|
|
|
|
|
my @arr = @{$avg{$key}}; |
|
0
|
|
|
|
|
|
|
143
|
0
|
|
|
|
|
|
my $arrLen = @arr; |
144
|
0
|
|
|
|
|
|
for(my $a = 0; $a < $arrLen; $a++){ |
145
|
0
|
|
|
|
|
|
my $wekaThing = $wekaAttr[$a]; |
146
|
0
|
|
|
|
|
|
my $entry = $arr[$a]; |
147
|
|
|
|
|
|
|
#try to align the values based on the length |
148
|
0
|
|
|
|
|
|
my $tab = ""; |
149
|
0
|
0
|
|
|
|
|
if(length($wekaThing) > 7){ |
|
|
0
|
|
|
|
|
|
150
|
0
|
|
|
|
|
|
$tab = "\t"; |
151
|
|
|
|
|
|
|
}elsif(length($wekaThing) < 4){ |
152
|
0
|
|
|
|
|
|
$tab = "\t\t\t"; |
153
|
|
|
|
|
|
|
}else{ |
154
|
0
|
|
|
|
|
|
$tab = "\t\t"; |
155
|
|
|
|
|
|
|
} |
156
|
|
|
|
|
|
|
|
157
|
0
|
|
|
|
|
|
my $in = "$wekaThing" . $tab . "-\t$entry"; |
158
|
0
|
|
|
|
|
|
$uniSub->print2File($WEKA_SAVE, $in); |
159
|
|
|
|
|
|
|
} |
160
|
0
|
|
|
|
|
|
$uniSub->print2File($WEKA_SAVE, ""); |
161
|
|
|
|
|
|
|
} |
162
|
0
|
|
|
|
|
|
$uniSub->print2File($WEKA_SAVE, "\n"); |
163
|
|
|
|
|
|
|
#print the individual |
164
|
0
|
|
|
|
|
|
foreach my $key (sort keys %ind){ |
165
|
|
|
|
|
|
|
#header |
166
|
0
|
|
|
|
|
|
$uniSub->print2File($WEKA_SAVE, "$key - INDIVIDUAL"); |
167
|
0
|
|
|
|
|
|
$uniSub->print2File($WEKA_SAVE, "==================\n"); |
168
|
0
|
|
|
|
|
|
$uniSub->print2FileNoLine($WEKA_SAVE, "\t\t\t"); |
169
|
0
|
|
|
|
|
|
foreach my $r(@wekaAttr){ |
170
|
0
|
|
|
|
|
|
my $sr = ""; |
171
|
0
|
0
|
|
|
|
|
if(length($r) >= 6){$sr = (substr($r, 0, 6) . ".");} |
|
0
|
0
|
|
|
|
|
|
172
|
0
|
|
|
|
|
|
elsif(length($r) < 4){$sr = "$r\t";} |
173
|
0
|
|
|
|
|
|
else{$sr = $r;} |
174
|
0
|
|
|
|
|
|
$uniSub->print2FileNoLine($WEKA_SAVE, "$sr\t"); |
175
|
|
|
|
|
|
|
} |
176
|
0
|
|
|
|
|
|
$uniSub->print2File($WEKA_SAVE, "\n-------------------------------------------------------------------------------"); |
177
|
|
|
|
|
|
|
|
178
|
|
|
|
|
|
|
#print the lines |
179
|
0
|
|
|
|
|
|
my @arr = @{$ind{$key}}; |
|
0
|
|
|
|
|
|
|
180
|
0
|
|
|
|
|
|
my $arrLen = @arr; |
181
|
|
|
|
|
|
|
|
182
|
0
|
|
|
|
|
|
for(my $a = 0; $a < $arrLen;$a++){ |
183
|
0
|
|
|
|
|
|
my $b = $a + 1; |
184
|
0
|
|
|
|
|
|
my $entry = $arr[$a]; |
185
|
|
|
|
|
|
|
#printColorDebug("on_red", "$name - $entry"); |
186
|
|
|
|
|
|
|
|
187
|
0
|
|
|
|
|
|
my $in = "BUCKET $b\t$entry"; |
188
|
0
|
|
|
|
|
|
$uniSub->print2File($WEKA_SAVE, $in); |
189
|
|
|
|
|
|
|
} |
190
|
|
|
|
|
|
|
|
191
|
0
|
|
|
|
|
|
$uniSub->print2File($WEKA_SAVE, ""); |
192
|
|
|
|
|
|
|
} |
193
|
|
|
|
|
|
|
|
194
|
0
|
|
|
|
|
|
close $WEKA_SAVE; |
195
|
|
|
|
|
|
|
|
196
|
|
|
|
|
|
|
} |
197
|
|
|
|
|
|
|
|
198
|
|
|
|
|
|
|
|
199
|
|
|
|
|
|
|
#average the weka accuracy datas |
200
|
|
|
|
|
|
|
# input : $name <-- name of the file with the weka data |
201
|
|
|
|
|
|
|
# output : %featAvg <-- hash of each feature set's averages (in array form aligned to the wekaAttr) |
202
|
|
|
|
|
|
|
sub averageWekaData{ |
203
|
0
|
|
|
0
|
0
|
|
my $name = shift; |
204
|
0
|
|
|
|
|
|
my %featAvg = (); |
205
|
|
|
|
|
|
|
|
206
|
0
|
|
|
|
|
|
foreach my $item (@sets){ |
207
|
0
|
|
|
|
|
|
my %data = (); |
208
|
|
|
|
|
|
|
|
209
|
0
|
|
|
|
|
|
foreach my $bucket(@allBuckets){ |
210
|
|
|
|
|
|
|
|
211
|
|
|
|
|
|
|
#import the wekaman |
212
|
0
|
|
|
|
|
|
my $file = "$program_dir/_WEKAS/$weka_dir/$name" . "_WEKA_DATA/$item/$name" . "_accuracy_$bucket"; |
213
|
0
|
0
|
|
|
|
|
open (WEKA, "$file") || die ("Unable to import $file from _WEKAS directory\n $!"); |
214
|
|
|
|
|
|
|
|
215
|
|
|
|
|
|
|
#get lines |
216
|
0
|
|
|
|
|
|
my @lines = ; |
217
|
0
|
|
|
|
|
|
foreach my $line(@lines){chomp($line)}; |
|
0
|
|
|
|
|
|
|
218
|
0
|
|
|
|
|
|
my $len = @lines; |
219
|
|
|
|
|
|
|
|
220
|
|
|
|
|
|
|
#$uniSub->printColorDebug("on_red", "YOUR FACE!"); |
221
|
|
|
|
|
|
|
|
222
|
|
|
|
|
|
|
#file doesn't have any contents for some reason |
223
|
0
|
0
|
|
|
|
|
if($len == 0){ |
224
|
0
|
|
|
|
|
|
$uniSub->printColorDebug("red", "**ERROR: NOTHING FOUND FOR $file!**\n"); |
225
|
0
|
|
|
|
|
|
my $n = @wekaAttr; |
226
|
0
|
|
|
|
|
|
my @values = (0) x $n; |
227
|
0
|
|
|
|
|
|
push(@{$data{$bucket}}, @values); |
|
0
|
|
|
|
|
|
|
228
|
0
|
|
|
|
|
|
next; |
229
|
|
|
|
|
|
|
#return %featAvg; |
230
|
|
|
|
|
|
|
} |
231
|
|
|
|
|
|
|
|
232
|
|
|
|
|
|
|
#get the rest of the array |
233
|
0
|
|
|
|
|
|
my $keyword = "=== Error on test data ==="; |
234
|
0
|
|
|
|
|
|
my $index = $uniSub->getIndexofLine($keyword, \@lines); |
235
|
0
|
|
|
|
|
|
my @result = @lines[$index..$len]; |
236
|
|
|
|
|
|
|
|
237
|
|
|
|
|
|
|
#grab the only stuff you need |
238
|
0
|
|
|
|
|
|
my $weightWord = " Yes"; |
239
|
0
|
|
|
|
|
|
my $weightIndex = $uniSub->getIndexofLine($weightWord, \@result); |
240
|
0
|
|
|
|
|
|
my $weightLine = $result[$weightIndex]; |
241
|
|
|
|
|
|
|
|
242
|
|
|
|
|
|
|
#split it uuuuuup |
243
|
0
|
|
|
|
|
|
my @values = split /\s+/, $weightLine; |
244
|
0
|
|
|
|
|
|
my $valLen = @values; |
245
|
0
|
|
|
|
|
|
my $valLen2 = $valLen - 2; |
246
|
0
|
|
|
|
|
|
@values = @values[1..$valLen2]; |
247
|
|
|
|
|
|
|
#printArr(", ", @values); |
248
|
|
|
|
|
|
|
|
249
|
|
|
|
|
|
|
#add to overall |
250
|
0
|
|
|
|
|
|
push(@{$data{$bucket}}, @values); |
|
0
|
|
|
|
|
|
|
251
|
|
|
|
|
|
|
} |
252
|
|
|
|
|
|
|
|
253
|
0
|
|
|
|
|
|
my @averages = (); |
254
|
|
|
|
|
|
|
#add all the stuffs |
255
|
0
|
|
|
|
|
|
foreach my $bucket (@allBuckets){ |
256
|
0
|
|
|
|
|
|
my @wekaSet = @{$data{$bucket}}; |
|
0
|
|
|
|
|
|
|
257
|
0
|
|
|
|
|
|
my $wekaLen = @wekaSet; |
258
|
0
|
|
|
|
|
|
for(my $e = 0; $e < $wekaLen; $e++){ |
259
|
0
|
|
|
|
|
|
my $entry = $wekaSet[$e]; |
260
|
0
|
0
|
0
|
|
|
|
if($entry ne "NaN" and $entry ne "?"){ #if value is unknown (from WEKA) |
261
|
0
|
|
|
|
|
|
$averages[$e] += $entry; |
262
|
|
|
|
|
|
|
} |
263
|
|
|
|
|
|
|
} |
264
|
|
|
|
|
|
|
} |
265
|
|
|
|
|
|
|
|
266
|
|
|
|
|
|
|
#printArr(", ", @avgLens); |
267
|
|
|
|
|
|
|
|
268
|
|
|
|
|
|
|
#divide them |
269
|
0
|
|
|
|
|
|
foreach my $tea (@averages){ |
270
|
0
|
|
|
|
|
|
$tea /= $bucketsNum; |
271
|
|
|
|
|
|
|
} |
272
|
|
|
|
|
|
|
|
273
|
|
|
|
|
|
|
#add the averages for the specific feature |
274
|
0
|
|
|
|
|
|
push(@{$featAvg{$item}}, @averages); |
|
0
|
|
|
|
|
|
|
275
|
|
|
|
|
|
|
} |
276
|
|
|
|
|
|
|
#exit; |
277
|
0
|
|
|
|
|
|
return %featAvg; |
278
|
|
|
|
|
|
|
} |
279
|
|
|
|
|
|
|
|
280
|
|
|
|
|
|
|
#retrieve the weighted average lines from the individual buckets |
281
|
|
|
|
|
|
|
# input : $name <-- name of the file with the weka data |
282
|
|
|
|
|
|
|
# output : %featData <-- hash of arrays for each line of individual data for the buckets |
283
|
|
|
|
|
|
|
sub individualWekaData{ |
284
|
0
|
|
|
0
|
0
|
|
my $name = shift; |
285
|
0
|
|
|
|
|
|
my %featData = (); |
286
|
|
|
|
|
|
|
|
287
|
0
|
|
|
|
|
|
foreach my $item (@sets){ #for each feature specified |
288
|
0
|
|
|
|
|
|
my @data = (); |
289
|
|
|
|
|
|
|
|
290
|
0
|
|
|
|
|
|
foreach my $bucket(@allBuckets){ #for each bucket in the set |
291
|
|
|
|
|
|
|
|
292
|
|
|
|
|
|
|
#import the wekaman |
293
|
0
|
|
|
|
|
|
my $file = "$program_dir/_WEKAS/$weka_dir/$name" . "_WEKA_DATA/$item/$name" . "_accuracy_$bucket"; |
294
|
0
|
0
|
|
|
|
|
open (WEKA, "$file") || die ("Unable to import $file from _WEKAS directory $!"); |
295
|
|
|
|
|
|
|
|
296
|
|
|
|
|
|
|
#get lines |
297
|
0
|
|
|
|
|
|
my @lines = ; |
298
|
0
|
|
|
|
|
|
foreach my $line(@lines){chomp($line)}; |
|
0
|
|
|
|
|
|
|
299
|
0
|
|
|
|
|
|
my $len = @lines; |
300
|
|
|
|
|
|
|
|
301
|
|
|
|
|
|
|
#if the contents of the file are empty for some reason |
302
|
0
|
0
|
|
|
|
|
if($len == 0){ |
303
|
0
|
|
|
|
|
|
$uniSub->printColorDebug("red", "**ERROR: NOTHING FOUND FOR $file!**\n"); |
304
|
0
|
|
|
|
|
|
my $n = @wekaAttr; |
305
|
0
|
|
|
|
|
|
my @values = (0) x $n; |
306
|
0
|
|
|
|
|
|
my $valueLine = join(" ", @values); |
307
|
0
|
|
|
|
|
|
push(@data, $valueLine); |
308
|
0
|
|
|
|
|
|
next; |
309
|
|
|
|
|
|
|
#return %featAvg; |
310
|
|
|
|
|
|
|
} |
311
|
|
|
|
|
|
|
|
312
|
|
|
|
|
|
|
#get the rest of the array |
313
|
0
|
|
|
|
|
|
my $keyword = "=== Error on test data ==="; |
314
|
0
|
|
|
|
|
|
my $index = $uniSub->getIndexofLine($keyword, \@lines); |
315
|
0
|
|
|
|
|
|
my @result = @lines[$index..$len]; |
316
|
|
|
|
|
|
|
|
317
|
|
|
|
|
|
|
#grab the only stuff you need |
318
|
0
|
|
|
|
|
|
my $weightWord = " Yes"; |
319
|
0
|
|
|
|
|
|
my $weightIndex = $uniSub->getIndexofLine($weightWord, \@result); |
320
|
0
|
|
|
|
|
|
my $weightLine = $result[$weightIndex]; |
321
|
|
|
|
|
|
|
|
322
|
|
|
|
|
|
|
#split it up |
323
|
0
|
|
|
|
|
|
my @values = split /\s+/, $weightLine; |
324
|
0
|
|
|
|
|
|
my $valLen = @values; |
325
|
0
|
|
|
|
|
|
$valLen -= 2; |
326
|
0
|
|
|
|
|
|
my @values2 = @values[1..$valLen]; |
327
|
|
|
|
|
|
|
|
328
|
|
|
|
|
|
|
#rejoin it |
329
|
0
|
|
|
|
|
|
my $valLine = join("\t", @values2); |
330
|
|
|
|
|
|
|
|
331
|
|
|
|
|
|
|
#add it |
332
|
0
|
|
|
|
|
|
push (@data, $valLine); |
333
|
|
|
|
|
|
|
} |
334
|
0
|
|
|
|
|
|
push (@{$featData{$item}}, @data); |
|
0
|
|
|
|
|
|
|
335
|
|
|
|
|
|
|
} |
336
|
0
|
|
|
|
|
|
return %featData; |
337
|
|
|
|
|
|
|
} |
338
|
|
|
|
|
|
|
|
339
|
|
|
|
|
|
|
# grabs the majority sense-label baseline for the file |
340
|
|
|
|
|
|
|
# input : $name <-- name of the file to get the entities from |
341
|
|
|
|
|
|
|
# output : $msb <-- # of no instances / # of yes instances |
342
|
|
|
|
|
|
|
sub getMSB{ |
343
|
0
|
|
|
0
|
0
|
|
my $file = shift; |
344
|
|
|
|
|
|
|
#get the name of the file |
345
|
0
|
|
|
|
|
|
my @n = split '/', $file; |
346
|
0
|
|
|
|
|
|
my $l = @n; |
347
|
0
|
|
|
|
|
|
my $filename = $n[$l - 1]; |
348
|
0
|
|
|
|
|
|
$filename = lc($filename); |
349
|
|
|
|
|
|
|
|
350
|
|
|
|
|
|
|
#import the data from the file |
351
|
0
|
0
|
|
|
|
|
if($program_dir ne ""){open (FILE, "$program_dir/$file") || die ("what is this '$program_dir/$file' you speak of?\n$!");} |
|
0
|
0
|
|
|
|
|
|
352
|
0
|
0
|
|
|
|
|
else{open (FILE, "$file") || die ("what is this '$file' you speak of? $!\n");} |
353
|
0
|
|
|
|
|
|
my @fileLines = ; |
354
|
0
|
|
|
|
|
|
foreach my $l(@fileLines){ |
355
|
0
|
|
|
|
|
|
$l = lc($l); |
356
|
|
|
|
|
|
|
} |
357
|
|
|
|
|
|
|
|
358
|
|
|
|
|
|
|
#clean it up for two separate sets |
359
|
0
|
|
|
|
|
|
my @tagSet = retagSet($filename, \@fileLines); |
360
|
|
|
|
|
|
|
|
361
|
|
|
|
|
|
|
#count and get results |
362
|
0
|
|
|
|
|
|
return countInst(\@tagSet); |
363
|
|
|
|
|
|
|
} |
364
|
|
|
|
|
|
|
|
365
|
|
|
|
|
|
|
# counts the # of no instances / # of all instances |
366
|
|
|
|
|
|
|
# input : @set <-- set of lines to read and count |
367
|
|
|
|
|
|
|
# output : $msb <-- # of no instances / # of all instances |
368
|
|
|
|
|
|
|
sub countInst{ |
369
|
0
|
|
|
0
|
0
|
|
my $set_ref = shift; |
370
|
0
|
|
|
|
|
|
my @set = @$set_ref; |
371
|
|
|
|
|
|
|
|
372
|
0
|
|
|
|
|
|
my $no_inst = 0; |
373
|
0
|
|
|
|
|
|
my $all_inst = 0; |
374
|
|
|
|
|
|
|
|
375
|
0
|
|
|
|
|
|
foreach my $line(@set){ |
376
|
0
|
|
|
|
|
|
my @words = split(" ", $line); |
377
|
0
|
|
|
|
|
|
foreach my $word(@words){ |
378
|
|
|
|
|
|
|
#print($word . "\n"); |
379
|
|
|
|
|
|
|
#if the word doesn't have a entity tag |
380
|
0
|
0
|
|
|
|
|
if(!($word =~ /([\s\S]*($entId)$)/)){ |
381
|
0
|
|
|
|
|
|
$no_inst++; #count the words not annotated |
382
|
|
|
|
|
|
|
} |
383
|
|
|
|
|
|
|
|
384
|
0
|
|
|
|
|
|
$all_inst++; #count all the words |
385
|
|
|
|
|
|
|
} |
386
|
|
|
|
|
|
|
} |
387
|
|
|
|
|
|
|
|
388
|
|
|
|
|
|
|
#return the ratio |
389
|
0
|
|
|
|
|
|
return ($no_inst / $all_inst); |
390
|
|
|
|
|
|
|
} |
391
|
|
|
|
|
|
|
|
392
|
|
|
|
|
|
|
|
393
|
|
|
|
|
|
|
|
394
|
|
|
|
|
|
|
###### RETAGS THE LINE ###### |
395
|
|
|
|
|
|
|
|
396
|
|
|
|
|
|
|
# turns the tagged entity words into special words with <> for the context words |
397
|
|
|
|
|
|
|
# input : $input <-- the line to retag |
398
|
|
|
|
|
|
|
# $id <-- the id within the tag to look for |
399
|
|
|
|
|
|
|
# output : (.arff files) |
400
|
|
|
|
|
|
|
sub retag{ |
401
|
0
|
|
|
0
|
0
|
|
my $input = shift; |
402
|
0
|
|
|
|
|
|
my $id = shift; |
403
|
|
|
|
|
|
|
|
404
|
0
|
|
|
|
|
|
$id = lc($id); |
405
|
0
|
|
|
|
|
|
my $line = lc($input); |
406
|
|
|
|
|
|
|
|
407
|
|
|
|
|
|
|
#get rid of any tags |
408
|
0
|
|
|
|
|
|
my @words = split (" ", $line); |
409
|
0
|
|
|
|
|
|
my @newSet = (); |
410
|
0
|
|
|
|
|
|
my $charact = 0; |
411
|
0
|
|
|
|
|
|
foreach my $word (@words){ |
412
|
0
|
0
|
|
|
|
|
if($charact){ #if currently annotating |
413
|
0
|
0
|
|
|
|
|
if($word eq ""){ #stop tagging annotations |
414
|
0
|
|
|
|
|
|
$charact = 0; |
415
|
|
|
|
|
|
|
}else{ #add the entity tagger to the words |
416
|
0
|
|
|
|
|
|
my $charWord = "$word"."$entId"; |
417
|
0
|
|
|
|
|
|
push @newSet, $charWord; |
418
|
|
|
|
|
|
|
} |
419
|
|
|
|
|
|
|
}else{ |
420
|
0
|
0
|
|
|
|
|
if($word eq ""){ #start annotating |
421
|
0
|
|
|
|
|
|
$charact = 1; |
422
|
|
|
|
|
|
|
}else{ #add normal word |
423
|
0
|
|
|
|
|
|
push @newSet, $word; |
424
|
|
|
|
|
|
|
} |
425
|
|
|
|
|
|
|
} |
426
|
|
|
|
|
|
|
} |
427
|
|
|
|
|
|
|
|
428
|
|
|
|
|
|
|
#clean up the new line |
429
|
0
|
|
|
|
|
|
my $new_line = join " ", @newSet; |
430
|
0
|
|
|
|
|
|
$new_line =~s/\b$entId\b//g; |
431
|
0
|
|
|
|
|
|
$new_line = $uniSub->cleanWords($new_line); |
432
|
0
|
|
|
|
|
|
return $new_line; |
433
|
|
|
|
|
|
|
} |
434
|
|
|
|
|
|
|
|
435
|
|
|
|
|
|
|
# turns the tagged entity words in the entire file into special words with <> for the context words |
436
|
|
|
|
|
|
|
# input : $name <-- the name of the file to use as the id tag |
437
|
|
|
|
|
|
|
# @lines <-- the set of lines to retag |
438
|
|
|
|
|
|
|
# output : @tagSet <-- set of retagged lines |
439
|
|
|
|
|
|
|
sub retagSet{ |
440
|
0
|
|
|
0
|
0
|
|
my $name = shift; |
441
|
0
|
|
|
|
|
|
my $lines_ref = shift; |
442
|
0
|
|
|
|
|
|
my @lines = @$lines_ref; |
443
|
|
|
|
|
|
|
|
444
|
0
|
|
|
|
|
|
my @tagSet = (); |
445
|
0
|
|
|
|
|
|
foreach my $line (@lines){ |
446
|
|
|
|
|
|
|
#retag the line |
447
|
0
|
|
|
|
|
|
chomp($line); |
448
|
0
|
|
|
|
|
|
my $tag_line = retag($line, $name); |
449
|
|
|
|
|
|
|
|
450
|
|
|
|
|
|
|
#add it to the set |
451
|
0
|
|
|
|
|
|
push @tagSet, $tag_line; |
452
|
|
|
|
|
|
|
} |
453
|
0
|
|
|
|
|
|
return @tagSet; |
454
|
|
|
|
|
|
|
} |
455
|
|
|
|
|
|
|
|
456
|
|
|
|
|
|
|
1; |