API tools faq
paste
Guest User

Deteksi Plagiarisme Dengan Winnowing

a guest
Nov 9th, 2015
1,463
0
Never
Add comment
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
PHP 9.48 KB | None | 0 0
raw download clone embed print report
  1. <?php
  2. $kalimat1 = isset($_POST['kalimat1']) ? $_POST['kalimat1'] : 'Indonesia Raya';
  3. $kalimat2 = isset($_POST['kalimat2']) ? $_POST['kalimat2'] : 'Indonesia Jaya';
  4. $n = isset($_POST['n']) ? $_POST['n'] : 5;
  5. $window = isset($_POST['window']) ? $_POST['window'] : 4;
  6. $prima = isset($_POST['prima']) ? $_POST['prima'] : 2;
  7.  
  8. ?>
  9. <html>
  10. <head>
  11. <title>Deteksi Plagiarisme Memanfaatkan Algoritma Winnowing</title>
  12. </head>
  13. <body>
  14. <form method="POST" action="winnowing.php">
  15. <p>Kalimat 1 : </p>
  16. <textarea name="kalimat1" style="width :100%" rows="5"><?php echo $kalimat1;?></textarea>
  17. <p>Kalimat 2 : </p>
  18. <textarea name="kalimat2" style="width :100%" rows="5"><?php echo $kalimat2;?></textarea>
  19. <p>N Gram : <input type ="text" name = "n" value = "<?php echo $n;?>" ></p>
  20. <p>Window : <input type ="text" name = "window" value = "<?php echo $window;?>" ></p>
  21. <p>Bilangan Prima :
  22. <select name = "prima">
  23. <option>Pilih Salah Satu</option>
  24. <?php
  25. for($i = 2; $i < 100; $i++){
  26.     $hitung = 0;
  27.     for($j = 1; $j <= $i; $j++){
  28.         if(($i % $j) == 0) $hitung++;
  29.     }
  30.     if($hitung == 2) {
  31.         $selected = ''; if($prima == $i) $selected = ' selected';
  32.         echo '<option value="'.$i.'" '.$selected.'>'.$i.'</option>';
  33.     }
  34. }
  35. ?>
  36. </select>
  37. </p>
  38. <p><input type=Submit value="Proses"></p>
  39. </form>
  40. <?php
  41. if(!$_POST) {
  42.     echo "</body>\n</html>";   
  43.     exit;
  44. }
  45.  
  46. $w = new winnowing($kalimat1, $kalimat2);
  47. $w->SetPrimeNumber($prima);
  48. $w->SetNGramValue($n);
  49. $w->SetNWindowValue($window);
  50.  
  51. $w->process();
  52.  
  53. ?>
  54. <h2>Hasil Proses</h2>
  55. <table style="width : 100%;">
  56. <tr>
  57. <td>N-GRAM Kalimat 1</td>
  58. <td>N-GRAM Kalimat 2</td>
  59. </tr>
  60. <tr>
  61. <td>
  62. <textarea style="width :100%; overflow-y:scroll;" rows="5">
  63. <?php
  64. $s ='';
  65. foreach($w->GetNGramFirst() as $ng){
  66.     $s .= $ng.' ';
  67. }
  68. echo rtrim($s, ' ');
  69. ?>
  70. </textarea>
  71. </td>
  72. <td>
  73. <textarea style="width :100%;overflow-y:scroll;" rows="5">
  74. <?php
  75. $s ='';
  76. foreach($w->GetNGramSecond() as $ng){
  77.     $s .= $ng.' ';
  78. }
  79. echo rtrim($s, ' ');
  80. ?>
  81. </textarea>
  82. </td>
  83. </tr>
  84. <tr>
  85. <td>Rolling Hash Kalimat 1</td>
  86. <td>Rolling Hash Kalimat 2</td>
  87. </tr>
  88. <tr>
  89. <td><textarea style="width :100%; overflow-y:scroll;" rows="5">
  90. <?php
  91. $s='';
  92. foreach($w->GetRollingHashFirst() as $rl){
  93.     $s .= $rl.' ';
  94. }
  95. echo rtrim($s, ' ');
  96. ?>
  97. </textarea></td>
  98. <td><textarea style="width :100%;overflow-y:scroll;" rows="5">
  99. <?php
  100. $s='';
  101. foreach($w->GetRollingHashSecond() as $rl){
  102.     $s .= $rl.' ';
  103. }
  104. echo rtrim($s, ' ');
  105. ?>
  106. </textarea></td>
  107. </tr>
  108. <tr>
  109. <td>Window Kalimat 1</td>
  110. <td>Window Kalimat 2</td>
  111. </tr>
  112. <tr>
  113. <td><textarea style="width :100%; overflow-y:scroll;" rows="5">
  114. <?php
  115. $wd = $w->GetWindowFirst();
  116. for($i = 0; $i< count($wd); $i++){
  117.     $s = '';
  118.     for($j=0; $j < $window; $j++){
  119.         $s .= $wd[$i][$j]. ' ';
  120.     }
  121.     echo "W-".($i+1)." : {".rtrim($s, ' ')."}\n";
  122. }
  123. ?>
  124. </textarea></td>
  125. <td><textarea style="width :100%;overflow-y:scroll;" rows="5">
  126. <?php
  127. $wd = $w->GetWindowSecond();
  128. for($i = 0; $i< count($wd); $i++){
  129.     $s = '';
  130.     for($j=0; $j < $window; $j++){
  131.         $s .= $wd[$i][$j]. ' ';
  132.     }
  133.     echo "W-".($i+1)." : {".rtrim($s, ' ')."}\n";
  134. }
  135. ?>
  136. </textarea></td>
  137. </tr>
  138. <tr>
  139. <td>Fingerprints Kalimat 1</td>
  140. <td>Fingerprints Kalimat 2</td>
  141. </tr>
  142. <tr>
  143. <td><textarea style="width :100%; overflow-y:scroll;" rows="5">
  144. <?php
  145. $s='';
  146. foreach($w->GetFingerprintsFirst() as $fp){
  147.     $s .= $fp.' ';
  148. }
  149. echo rtrim($s, ' ');
  150. ?>
  151. </textarea></td>
  152. <td><textarea style="width :100%;overflow-y:scroll;" rows="5">
  153. <?php
  154. $s='';
  155. foreach($w->GetFingerprintsSecond() as $fp){
  156.     $s .= $fp.' ';
  157. }
  158. echo rtrim($s, ' ');
  159.  
  160. $count_fingers1 = count($w->GetFingerprintsFirst());
  161. $count_fingers2 = count($w->GetFingerprintsSecond());
  162.  
  163. $count_union_fingers = count(array_merge($w->GetFingerprintsFirst(), $w->GetFingerprintsSecond()));
  164. $count_intersect_fingers = count(array_intersect($w->GetFingerprintsFirst(), $w->GetFingerprintsSecond()));
  165.  
  166. ?>
  167. </textarea>
  168. </td>
  169. </tr>
  170. </table>
  171. <p>Jumlah Fingerprints kalimat 1 = <?php echo $count_fingers1;?></p>
  172. <p>Jumlah Fingerprints kalimat 2 - <?php echo $count_fingers2;?></p>
  173. <p>Union (Gabungan) Fingerprints 1 dan 2 = <?php echo $count_union_fingers;?></p>
  174. <p>Intersection (fingerprints yang sama) = <?php echo $count_intersect_fingers;?></p>
  175. <p>(Union - Intersection) = <?php echo ($count_union_fingers - $count_intersect_fingers);?></p>
  176. <p>Prosentase Plagiarisme </p>
  177. <p>&nbsp;&nbsp;Koefisien Jaccard = (Intersection / (Union-Intersection)) * 100 </p>
  178. <p>&nbsp;&nbsp;(<?php echo $count_intersect_fingers ."/".($count_union_fingers - $count_intersect_fingers).") * 100 = ".$w->GetJaccardCoefficient();?> %</p>
  179. </body>
  180. </html>
  181.  
  182. <?php
  183. class winnowing {
  184.     private $word1 = '';
  185.     private $word2 = '';
  186.  
  187.     //input properties
  188.     private $prime_number = 3;
  189.     private $n_gram_value = 2;
  190.     private $n_window_value = 4;    
  191.    
  192.     //output properties
  193.     private $arr_n_gram1;
  194.     private $arr_n_gram2;
  195.     private $arr_rolling_hash1;
  196.     private $arr_rolling_hash2;
  197.     private $arr_window1;
  198.     private $arr_window2;
  199.     private $arr_fingerprints1;
  200.     private $arr_fingerprints2;
  201.  
  202.     public function SetPrimeNumber($value){
  203.         $this->prime_number = $value;
  204.     }  
  205.     public function SetNGramValue($value){
  206.         $this->n_gram_value = $value;
  207.     }
  208.     public function SetNWindowValue($value){
  209.         $this->n_window_value = $value;
  210.     }
  211.     public function GetNGramFirst(){
  212.         return $this->arr_n_gram1;
  213.     }
  214.     public function GetNGramSecond(){
  215.         return $this->arr_n_gram2;
  216.     }
  217.     public function GetRollingHashFirst(){
  218.         return $this->arr_rolling_hash1;
  219.     }
  220.     public function GetRollingHashSecond(){
  221.         return $this->arr_rolling_hash2;
  222.     }
  223.     public function GetWindowFirst(){
  224.         return $this->arr_window1;
  225.     }
  226.     public function GetWindowSecond(){
  227.         return $this->arr_window2;
  228.     }
  229.     public function GetFingerprintsFirst(){
  230.         return $this->arr_fingerprints1;
  231.     }
  232.     public function GetFingerprintsSecond(){
  233.         return $this->arr_fingerprints2;
  234.     }
  235.     public function GetJaccardCoefficient($prosen = true){
  236.         if($prosen)
  237.             return round( ($this->jaccard_coefficient * 100), 2);
  238.         else
  239.             return $this->jaccard_coefficient;
  240.     }
  241.  
  242.     function __construct($w1, $w2){
  243.         $this->word1 = $w1;
  244.         $this->word2 = $w2;
  245.     }
  246.  
  247.     public function process(){
  248.         if (($this->word1 == '') || ($this->word2 == '')) exit;
  249.  
  250.         //langkah 1 : buang semua huruf yang bukan kelompok [a-z A-Z 0-9] dan ubah menjadi huruf kecil semua (lowercase)
  251.         $this->word1 = strtolower(str_replace(' ','',preg_replace("/[^a-zA-Z0-9\s-]/", "", $this->word1)));
  252.         $this->word2 = strtolower(str_replace(' ','',preg_replace("/[^a-zA-Z0-9\s-]/", "", $this->word2)));
  253.  
  254.         //langkah 2 : buat N-Gram
  255.         $this->arr_n_gram1 = $this->n_gram($this->word1, $this->n_gram_value);
  256.         $this->arr_n_gram2 = $this->n_gram($this->word2, $this->n_gram_value);
  257.  
  258.         //langkah 3 : rolling hash untuk masing-masing n gram
  259.         $this->arr_rolling_hash1 = $this->rolling_hash($this->arr_n_gram1);
  260.         $this->arr_rolling_hash2 = $this->rolling_hash($this->arr_n_gram2);
  261.  
  262.         //langkah 4 : buat windowing untuk masing-masing tabel hash
  263.         $this->arr_window1 = $this->windowing($this->arr_rolling_hash1, $this->n_window_value);
  264.         $this->arr_window2 = $this->windowing($this->arr_rolling_hash2, $this->n_window_value);
  265.  
  266.         //langkah 5 : cari nilai minimum masing-masing window table (fingerprints)
  267.         $this->arr_fingerprints1 = $this->fingerprints($this->arr_window1);
  268.         $this->arr_fingerprints2 = $this->fingerprints($this->arr_window2);
  269.  
  270.         //langkah 6 : hitung koefisien plagiarisme memanfaatkan persamaan Jaccard Coefficient
  271.         $this->jaccard_coefficient = $this->jaccard_coefficient($this->arr_fingerprints1, $this->arr_fingerprints2);
  272.     }
  273.  
  274.     private function n_gram($word, $n) {
  275.         $ngrams = array();
  276.         $length = strlen($word);
  277.         for($i = 0; $i < $length; $i++) {
  278.                 if($i > ($n - 2)) {
  279.                         $ng = '';
  280.                         for($j = $n-1; $j >= 0; $j--) {
  281.                                 $ng .= $word[$i-$j];
  282.                         }
  283.                         $ngrams[] = $ng;
  284.                 }
  285.         }
  286.         return $ngrams;
  287.     }
  288.  
  289.     private function char2hash($string) {
  290.         if (strlen($string) == 1) {
  291.             return ord($string);
  292.         } else {
  293.             $result = 0;
  294.             $length = strlen($string);
  295.             for ($i = 0; $i < $length; $i++) {
  296.                 $result += ord(substr($string, $i, 1)) * pow($this->prime_number, $length-$i);
  297.             }
  298.             return $result;
  299.         }
  300.     }
  301.  
  302.     private function rolling_hash($ngram){
  303.         $roll_hash = array();  
  304.         foreach($ngram as $ng){
  305.             $roll_hash[] = $this->char2hash($ng);
  306.         }
  307.         return $roll_hash;
  308.     }
  309.  
  310.     private function windowing($rolling_hash, $n){
  311.         $ngram = array();
  312.         $length = count($rolling_hash);
  313.         $x = 0;
  314.         for($i = 0; $i < $length; $i++){
  315.             if($i > ($n - 2)) {
  316.                 $ngram[$x] = array();
  317.                 $y = 0;
  318.                 for($j = $n-1; $j >= 0; $j--){
  319.                     $ngram[$x][$y] = $rolling_hash[$i-$j];
  320.                     $y++;
  321.                 }
  322.                 $x++;
  323.             }
  324.         }
  325.         //echo $x.' '.$y;
  326.         return $ngram;
  327.     }
  328.  
  329.     private function fingerprints($window_table){
  330.         $fingers = array();
  331.         for($i = 0; $i < count($window_table); $i++){
  332.             $min = $window_table[$i][0];
  333.             for($j = 1 ; $j < $this->n_window_value; $j++){
  334.                 if($min > $window_table[$i][$j])
  335.                     $min = $window_table[$i][$j];
  336.             }
  337.             $fingers[] = $min;
  338.         }
  339.         return $fingers;   
  340.     }
  341.  
  342.     private function jaccard_coefficient($fingerprint1, $fingerprint2){
  343.         $arr_intersect = array_intersect( $fingerprint1, $fingerprint2 );
  344.         $arr_union = array_merge( $fingerprint1, $fingerprint2 );
  345.    
  346.         $count_intersect_fingers = count($arr_intersect);
  347.         $count_union_fingers = count( $arr_union );
  348.  
  349.         $coefficient = $count_intersect_fingers /
  350.             ($count_union_fingers - $count_intersect_fingers );
  351.        
  352.         return $coefficient;
  353.     }
  354. }
  355. ?>
Advertisement
Add Comment
Please, Sign In to add comment
Public Pastes
We use cookies for various purposes including analytics. By continuing to use Pastebin, you agree to our use of cookies as described in the Cookies Policy.  OK, I Understand
Not a member of Pastebin yet?
Sign Up, it unlocks many cool features!