Automatically Generate Tags

Author:

• Zhe Yu (zyu9)

Structure

data

• data folder contains the raw data of this project, namely anime.txt.

dump

• dump folder contains the results in the form of pickle.

fig

• fig folder contains all the figures drawn to visualize the results.

reports

• reports folder contains proposal, midway milestone, poster, and final report of the project.

src

• src folder contains all the source code.

Instruction of running the code

• Run GeTagAnime.py to get all the classification result and F scores, the results will be dumped into pickle files in dump folder.
• Run draw.py to generate figures from the dumped results. Figures will be saved in fig folder.

Libraries needed before run

• pdb

• numpy

• matplotlib

• scikit-learn : conda install scikit-learn in the shell

• pickle

• nltk :

  conda install nltk
  
  python
  
  >>> import nltk
  
  >>> nltk.download()
  

  Then choose stopwords to install.

• have fun!!