tuned better

app.py

@@ -9,9 +9,9 @@ import nltk
 nltk.download('all')
 import string
 from sklearn.feature_extraction.text import TfidfVectorizer
+import random
 # import fastai
 
-
 def similarity(input, joke):
     return cosine_similarity(input, joke)
 
@@ -74,21 +74,22 @@ def LemNormalize(text):
   return LemTokens(nltk.word_tokenize(text.lower().translate(remove_punct_dict)))
 
 def NLTK(input):
-    f = open('corpus.txt', errors='strict')
+    f = open('/content/corpus.txt', errors='strict')
     data = f.read()
     data = data.lower()
+    data = data + input.lower()
     sent_tokens = nltk.sent_tokenize(data)
-    return bot(sent_tokens)
+    return bot(sent_tokens, input)
 
-def bot(sent_tokens):
+def bot(sent_tokens, input):
   robo1_response = ''
   TfidfVec = TfidfVectorizer(tokenizer = LemNormalize, stop_words='english')
   tfidf = TfidfVec.fit_transform(sent_tokens)
   vals = cosine_similarity(tfidf[-1], tfidf)
-  idx = vals.argsort()[0][-2]
+  idx = random.randint(0, len(vals.argsort()[0]))
   flat = vals.flatten()
   flat.sort()
-  req_tfidf = flat[-2]
+  req_tfidf = flat[-5]
   if (req_tfidf == 0):
     robo1_response= robo1_response+"I could not answer this right now but you can contact the head of our dept (PUSPHA RAJ)." # add the dept recommendation engine and contact details
     return robo1_response