Is there any good open-source or freely available Chinese segmentation algorithm available?

Sebastian picture Sebastian · Apr 29, 2011 · Viewed 9.7k times · Source

As phrased in the question, I'm looking for a free and/or open-source text-segmentation algorithm for Chinese, I do understand it is a very difficult task to solve, as there are many ambiguities involed. I know there's google's API, but well it is rather a black-box, i.e. not many information of what it is doing are passing through.

Answer

lschin picture lschin · May 19, 2011

The keyword text-segmentation for Chinese should be 中文分词 in Chinese.

Good and active open-source text-segmentation algorithm :

  1. 盘古分词(Pan Gu Segment) : C#, Snapshot
  2. ik-analyzer : Java
  3. ICTCLAS : C/C++, Java, C#, Demo
  4. NlpBamboo : C, PHP, PostgreSQL
  5. HTTPCWS : based on ICTCLAS, Demo
  6. mmseg4j : Java
  7. fudannlp : Java, Demo
  8. smallseg : Python, Java, Demo
  9. nseg : NodeJS
  10. mini-segmenter: python

Other

  1. Google Code : http://code.google.com/query/#q=中文分词
  2. OSChina (Open Source China)

Sample

  1. Google Chrome (Chromium) : src, cc_cedict.txt (73,145 Chinese words/pharases)

    • In text field or textarea of Google Chrome with Chinese sentences, press Ctrl+ or Ctrl+

    • Double click on 中文分词指的是将一个汉字序列切分成一个一个单独的词