A python package to predict the Class of a Question i.e. whether Who , What , When , Affirmation or Unknown.
Required packages (for Python 2.7.6)
joblib==0.9.4
nltk==3.2
numpy==1.10.4
pandas==0.17.0
scikit-learn==0.17
scipy==0.17.0
sklearn==0.0python QClass.py "What is your name?"
WHAT
from QClass import getQClass
queslist=[
'Define metamorphosis',
'What was the date of the match?',
'Did John enjoy the match?',
'Who invented electricity?',
"Did you hear John's decided to go to business school?",
'Explain centrifugal force',
'What is the name of the god of water?',
'Is it dry outside?',
]
for ques in queslist:
print ques+"\t"+getQClass(ques) Define metamorphosis UNKNOWN
What was the date of the match ? WHEN
Did John enjoy the match ? Affirmation
Who invented electricity ? WHO
Did you hear John's decided to go to business school ? Affirmation
Explain centrifugal force UNKNOWN
What is the name of the god of water ? WHAT
Is it dry outside ? Affirmation
Place your train.csv in the data folder in the following tab-separated format
Question Label
Note: The Affirmation cases do not require training (they are heuristics based) and so should be included in the training file .
Run python training.py and the model would be saved in the model directory.