code for Natural Language Processing. (especially Korean.)

only using existing library, not developing any existing algorithm.

simple code for using Korean POS tagger (KoNLPy)

make basic corpus

• cbt4,5_chatall.csv
  • result : csv file
  • header : ID, corpus
• cbt5_chatall3000.csv
  • All same, but save only 3000 letters
• 0_0_type.csv
  • cbt, week, type all corpus

Labeling user.

Using pos analyzed data, select feature and make BagofWords. Then, Classify.

Calculate tf-idf

plot scatter graph and box chart.

do dbscan, sort, join

do Self-Organizing Maps