def AnalyzeResultforQuestionWithIDF(utterence,no):
y=wordvec.vectorize_query(utterence,IDFset)
scores=[]
query=Data_Cleaner(utterence)
processedQuestion=[]
for i in range(Question_vectors.__len__()):
x = np.array(Question_vectors[i])
score = spatial.distance.cosine(x, y)
processedQuestion.append(Data_Cleaner(TrainingSet.Question[i]))
if math.isnan(score):
scores.append(1)
else:
scores.append(score)
TestSet = pd.DataFrame(list(zip(TrainingSet.Question,processedQuestion, scores)),columns=['Question','Processed Question', 'Score'])
utterence=re.sub('[^A-Za-z0-9 ]+', '', utterence)
excel_name=str('_'.join(utterence.split())) +".xlsx"
excel_name="TestSet"+no+".xlsx"
TestSet.to_excel(excel_name)
index=scores.index(np.min(scores))
print("Index",index)
answer=TrainingSet.Answer[index]
question=TrainingSet.Question[index]
return query,question,answer,np.min(scores)
def getAnswer(utterence):
y = wordvec.vectorize_query(utterence, IDFset)
scores = []
for i in range(Question_vectors.__len__()):
x = np.array(Question_vectors[i])
score = spatial.distance.euclidean(x, y)
if math.isnan(score):
scores.append(1)
else:
scores.append(score)
index = scores.index(np.min(scores))
print("Index", index)
answer = TrainingSet.Answer[index]
question = TrainingSet.Question[index]
return question, answer