Update bm25_retreive_question.py

bm25_retreive_question.py

@@ -1,12 +1,19 @@
 from rank_bm25 import BM25Okapi
 import nltk
 from nltk.tokenize import word_tokenize
+import numpy as np
+import streamlit as st
+import random
+
+
 
 # Download NLTK data for tokenization
 nltk.download('punkt')
 
 class QuestionRetriever:
     def __init__(self):
+        self.question_history = []
+
         self.depression_questions = [
             "How often have you felt persistently low in mood or sad for most of the day?",
             "How often have you lost interest or pleasure in activities you used to enjoy?",
@@ -122,11 +129,33 @@ class QuestionRetriever:
         doc_scores = bm25.get_scores(tokenized_query)
 
         # Get the index of the most relevant document
-        most_relevant_doc_index = doc_scores.argmax()
+        # most_relevant_doc_index = doc_scores.argmax()
+
+        # Get the indices of documents sorted by score in descending order
+        sorted_indices = np.argsort(doc_scores)[::-1]
+
+       # Flag to keep track of whether a relevant document index is found
+        found_new_question = False
+        self.question_history = st.session_state.asked_questions
+        print(self.question_history)
+
+        for i in range(len(sorted_indices)):
+            print(f"Checking {i+1} question - Index {sorted_indices[i]}th - {knowledge_base[sorted_indices[i]]}")
+            most_relevant_doc_index = sorted_indices[i]
+            if knowledge_base[most_relevant_doc_index] not in self.question_history:
+                found_new_question = True
+                question = knowledge_base[most_relevant_doc_index]
+                st.session_state.asked_questions.append(question)
+                print(f">>> Returning Top #{i+1} question")
+                return question
+
+        # If no relevant document index is found, return None or handle the case accordingly
+        if not found_new_question:
+            question = random.choice(knowledge_base)
 
         # Fetch the corresponding response from the knowledge base
-        response = knowledge_base[most_relevant_doc_index]
-        return response
+        # response = knowledge_base[most_relevant_doc_index]
+        return question
 
 if __name__ == "__main__":
     knowledge_base = depression_questions