Free language processing service and NLP C# code
Free Part-of-Speech Tagging Service Viterbi Part-of-Speech Tagger

Upload your file

  1. Give your text by typing, or copying & pasting, or file uploading.
  2. Check your options.
  3. Click 'Tag It'.

Additional Tagging Services:
  1. Build a tagger of any language
  2. Tag your corpora of any size

Conact me for details

The tagger you see here is completed in two steps. A probability-based tagger was first built out of pre-tagged corpora, employing Bayes' theorem, a Markov assumption, and the Viterbi algorithm. The tagger then went through a reinforcement machine learning process to improve its accuracy. When doing the actual tagging, the contextual transformation rules extracted from the training data will be applied to achieve the maximal accuracy.

    The tagger uses the Penn Treebank tagset. It's accuracy is about 97.1% on ordinary texts. Its speed, on a computer with 2G RAM and two 3.00 GHz Pentium duo processors, is about 26,000 words per second by a desktop application.

    For more information about this tagger, see Jiayun Han's MS thesis on part-of-speech tagging.

Skip Navigation LinksHome > Free NLP Online Services > Part-of-speech tagging