def predict_question(input_question):
'''
Runs through variable extraction and the question classifier to
predict the intended question.
Args: input_question (string) - user input question to answer
Return: return_tuple (tuple) - contains the user's input question,
the variable extracted input question,
the entity extracted, and the predicted
answer
'''
variable_extraction = Variable_Extraction()
entity, normalized_sentence = variable_extraction.\
extract_variables(input_question)
classifier = QuestionClassifier()
classifier.load_latest_classifier()
answer = classifier.classify_question(normalized_sentence)
return_tuple = (input_question, normalized_sentence,
entity, answer)
return return_tuple
def eval_model_accuracy():
question_generator = Generate_Data.Gen_Data()
generated_data, _ = question_generator.generate()
classifier = QuestionClassifier('question_set_clean.csv', use_new=True)
total = 0
samples = 0
for data in generated_data:
expect_question = data[0]
for i in range(1, len(data)):
samples += 1
test = data[i]
classified_question = classifier.classify_question(test)
if classified_question != expect_question:
print("--------")
print("Question: {}".format(test))
print(
"Expected to be classified as: {}".format(expect_question))
print("Got classification: {}".format(classified_question))
print("-------")
else:
total += 1
print("Accuracy: {}".format(float(total) / samples))
print(generated_data)
