syitilv / POSTagger-Indonesia

A POSTagger for Bahasa Indonesia

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

The POSTagger is using baseline method, decision tree (classification-based approach), and HMM-Viterbi (sequence-based approach).The results are baseline method gets the highest accuracy other than method which is 87%. Even though, Viterbi only differs 1% lower than baseline method and the classification method gets only 79%. Viterbi only differs by 1% because it also counts the maximum weight of the transition, rather than baseline that only use the maximum frequency of the chosen word to be tagged. This proves that baseline or Viterbi method are quite reliable to be used for POSTagging.

About

A POSTagger for Bahasa Indonesia


Languages

Language:Jupyter Notebook 100.0%