Update app.py

app.py

@@ -0,0 +1,118 @@
+# Install ipywidgets (if not already installed)
+!pip install ipywidgets
+
+# Imports
+import pandas as pd
+import ipywidgets as widgets
+from IPython.display import display
+
+# Load your CSV file
+df = pd.read_csv('mcq_dataset.csv')
+
+# Unique domains
+domains = df['domain'].unique().tolist()
+
+# Create dropdown widgets
+domain_dropdown = widgets.Dropdown(
+    options=domains,
+    description='Domain:',
+    style={'description_width': 'initial'}
+)
+
+subdomain_dropdown = widgets.Dropdown(
+    options=[],
+    description='Subdomain:',
+    style={'description_width': 'initial'}
+)
+
+# Update subdomains based on selected domain
+def update_subdomains(*args):
+    selected_domain = domain_dropdown.value
+    filtered = df[df['domain'] == selected_domain]
+    subdomains = filtered['subdomain'].unique().tolist()
+    subdomain_dropdown.options = subdomains
+
+domain_dropdown.observe(update_subdomains, 'value')
+update_subdomains()  # initial call
+
+display(domain_dropdown)
+display(subdomain_dropdown)
+
+from sklearn.feature_extraction.text import TfidfVectorizer
+from sklearn.metrics.pairwise import cosine_similarity
+
+def get_top_mcqs(user_input, domain, subdomain, df, top_n=10):
+    # Filter dataset by domain and subdomain
+    filtered_df = df[(df['domain'] == domain) & (df['subdomain'] == subdomain)]
+
+    if filtered_df.empty:
+        print("No MCQs found for the selected domain/subdomain.")
+        return pd.DataFrame()
+
+    # Combine user input with keywords for similarity comparison
+    documents = filtered_df['keywords'].tolist()
+    documents.insert(0, user_input)  # user input is first
+
+    # TF-IDF vectorization
+    vectorizer = TfidfVectorizer()
+    tfidf_matrix = vectorizer.fit_transform(documents)
+
+    # Compute cosine similarity between user input and all MCQs
+    cosine_sim = cosine_similarity(tfidf_matrix[0:1], tfidf_matrix[1:]).flatten()
+
+    # Get indices of top N most similar questions
+    top_indices = cosine_sim.argsort()[-top_n:][::-1]
+
+    # Retrieve top questions
+    top_questions = filtered_df.iloc[top_indices].copy()
+    top_questions['similarity_score'] = cosine_sim[top_indices]
+
+    return top_questions.reset_index(drop=True)
+
+
+# Assume the user selected values from the dropdowns
+selected_domain = domain_dropdown.value
+selected_subdomain = subdomain_dropdown.value
+
+# User inputs a topic or keywords
+user_input = input("Enter keywords or topic to focus on: ")
+
+# Retrieve top 10 relevant MCQs
+top_mcqs = get_top_mcqs(user_input, selected_domain, selected_subdomain, df)
+
+# Display them
+top_mcqs[['question', 'option1', 'option2', 'option3', 'option4', 'correct_answer', 'similarity_score']]
+
+
+def run_quiz(mcq_df):
+    score = 0
+
+    print("\n--- Starting the Quiz ---\n")
+
+    for i, row in mcq_df.iterrows():
+        print(f"Q{i+1}: {row['question']}")
+        print(f"A. {row['option1']}")
+        print(f"B. {row['option2']}")
+        print(f"C. {row['option3']}")
+        print(f"D. {row['option4']}")
+
+        answer_map = {
+            'A': row['option1'],
+            'B': row['option2'],
+            'C': row['option3'],
+            'D': row['option4']
+        }
+
+        user_ans = input("Your answer (A/B/C/D): ").strip().upper()
+        selected_option = answer_map.get(user_ans, None)
+
+        if selected_option == row['correct_answer']:
+            print("✅ Correct!\n")
+            score += 1
+        else:
+            print(f"❌ Incorrect! Correct answer: {row['correct_answer']}\n")
+
+    print(f"🎯 Quiz Complete! Your Score: {score} / 10")
+    return score
+
+user_score = run_quiz(top_mcqs)