0.33 版

README.md

@@ -253,6 +253,7 @@ array(18) {
 * JiebaAnalyse::extractTags($content, $top_k)
 * content 為待提取的文本
 * top_k 為返回幾個 TF/IDF 權重最大的關鍵詞，默認值為 20
+* 可使用 setStopWords 增加自定義 stop words
 
 代碼示例 (關鍵詞提取)
 
@@ -276,32 +277,60 @@ $content = file_get_contents("/path/to/your/dict/lyric.txt", "r");
 
 $tags = JiebaAnalyse::extractTags($content, $top_k);
 
+var_dump($tags);
+
+JiebaAnalyse::setStopWords('/path/to/your/dict/stop_words.txt');
+
+$tags = JiebaAnalyse::extractTags($content, $top_k);
+
 var_dump($tags);
 ```
 
 Output:
 ```php
 array(10) {
-  ["是否"]=>
-  float(1.2196321889395)
-  ["一般"]=>
-  float(1.0032459890209)
-  ["肌迫"]=>
-  float(0.64654314660465)
-  ["怯懦"]=>
-  float(0.44762844339349)
-  ["藉口"]=>
-  float(0.32327157330233)
-  ["逼不得已"]=>
-  float(0.32327157330233)
-  ["不安全感"]=>
-  float(0.26548304656279)
-  ["同感"]=>
-  float(0.23929673812326)
-  ["有把握"]=>
-  float(0.21043366018744)
-  ["空洞"]=>
-  float(0.20598261709442)
+  '沒有' =>
+  double(1.0592831964595)
+  '所謂' =>
+  double(0.90795702553671)
+  '是否' =>
+  double(0.66385043195443)
+  '一般' =>
+  double(0.54607060161899)
+  '雖然' =>
+  double(0.30265234184557)
+  '來說' =>
+  double(0.30265234184557)
+  '肌迫' =>
+  double(0.30265234184557)
+  '退縮' =>
+  double(0.30265234184557)
+  '矯作' =>
+  double(0.30265234184557)
+  '怯懦' =>
+  double(0.24364586159392)
+}
+array(10) {
+  '所謂' =>
+  double(1.1569129841516)
+  '一般' =>
+  double(0.69579963754677)
+  '矯作' =>
+  double(0.38563766138387)
+  '來說' =>
+  double(0.38563766138387)
+  '退縮' =>
+  double(0.38563766138387)
+  '雖然' =>
+  double(0.38563766138387)
+  '肌迫' =>
+  double(0.38563766138387)
+  '怯懦' =>
+  double(0.31045198493419)
+  '隨便說說' =>
+  double(0.19281883069194)
+  '一場' =>
+  double(0.19281883069194)
 }
 ```
 

composer.json

@@ -4,7 +4,7 @@
     "description": "結巴中文分詞（PHP 版本）：做最好的 PHP 中文分詞、中文斷詞組件",
     "keywords": ["Jieba", "PHP"],
     "license": "MIT",
-    "version": "0.32",
+    "version": "0.33",
     "authors": [
         {
             "name": "fukuball",

src/class/JiebaAnalyse.php

@@ -49,11 +49,17 @@ public static function init($options = array())
     {
 
         $defaults = array(
-            'mode'=>'default'
+            'mode'=>'default',
+            'dict'=>'normal'
         );
 
         $options = array_merge($defaults, $options);
 
+        if ($options['dict']=='big') {
+            $f_name = "idf.big.txt";
+        } else {
+            $f_name = "idf.txt";
+        }
         $content = fopen(dirname(dirname(__FILE__))."/dict/idf.txt", "r");
 
         while (($line = fgets($content)) !== false) {
@@ -72,6 +78,27 @@ public static function init($options = array())
         self::$median_idf = self::$idf_freq[$middle_key];
     }// end function init
 
+    /**
+     * Static method setStopWords
+     *
+     * @param string  $stop_words_path
+     * @param array   $options
+     *
+     * @return array $tags
+     */
+    public static function setStopWords($stop_words_path, $options = array())
+    {
+        $content = fopen($stop_words_path, "r");
+
+        while (($line = fgets($content)) !== false) {
+            $stop_word = strtolower(trim($line));
+            if (! in_array($stop_word, self::$stop_words)) {
+                array_push(self::$stop_words, $stop_word);
+            }
+        }
+        fclose($content);
+    }
+
     /**
      * Static method extractTags
      *

src/cmd/demo_extract_tags.php

@@ -24,12 +24,16 @@
 use Fukuball\Jieba\JiebaAnalyse;
 Jieba::init(array('mode'=>'test','dict'=>'big'));
 Finalseg::init();
-JiebaAnalyse::init();
+JiebaAnalyse::init(array('dict'=>'big'));
 
 $top_k = 10;
 $content = file_get_contents(dirname(dirname(__FILE__))."/dict/lyric.txt", "r");
 
 $tags = JiebaAnalyse::extractTags($content, $top_k);
+var_dump($tags);
+
+JiebaAnalyse::setStopWords(dirname(dirname(__FILE__)).'/dict/stop_words.txt');
 
+$tags = JiebaAnalyse::extractTags($content, $top_k);
 var_dump($tags);
 ?>

src/dict/stop_words.txt

@@ -0,0 +1,51 @@
+the
+of
+is
+and
+to
+in
+that
+we
+for
+an
+are
+by
+be
+as
+on
+with
+can
+if
+from
+which
+you
+it
+this
+then
+at
+have
+all
+not
+one
+has
+or
+that
+的
+了
+和
+是
+就
+都
+而
+及
+與
+著
+或
+一個
+沒有
+我們
+你們
+妳們
+他們
+她們
+是否