dialogue project (2014 spring)

Pipeline for conversation strategy

text input-output system
baseline/Interface.py

• NLU component
  code: baseline/Understander.py
  

• Retrival component
  code: basedline/Retrieval.py
  
  (reverse index can be used when database increased)
  

  The conversation databased we collected is here:
  https://drive.google.com/file/d/0B7XJjU2\_aDgsTnVZM0FjbVk0UTQ/edit
  

• Dialogue Management component
  code: baseline/Control.py
  

• NLG coponent
  code: baseline/NLG.py
  
  The templates and topic words are in baseline/resouce
  

• Using Google API
  Python Wrapper for jar pacakge based on Google API, testing the result from Google API
  
  code: analysis/Google.py
  

Algorithm details can be refered to the slides

Dependencies
nltk http://www.nltk.org/

Tkinter https://wiki.python.org/moin/TkInter