Skip to content

Latest commit

 

History

History
16 lines (14 loc) · 738 Bytes

README.md

File metadata and controls

16 lines (14 loc) · 738 Bytes
Raw

HMM-Viterbi

POS Tagging simulator using HMM method with Viterbi algorithm. Hidden Markov Model is one of the method to find the best sequence from a hidden path. This method could be use as a POS Tagger. Combined with Viterbi algorithm to reduce time complexity. Use Laplacian Smoothing to rise the accuracy.

How to run

To run the program on your terminal:

javac HMMViterbi.java
java HMMViterbi

The data

  • The training data should be put on a folder named Training. It could be more than one file.
  • The testing data should be put on a folder named Testing2. It could be more than one file.
  • The testing data should include the real tag for computing the accuracy.
  • The output will be put on a folder named Output