Update app.py

app.py

@@ -9,26 +9,27 @@ from sklearn.metrics.pairwise import cosine_similarity
 # ==========================================
 # 1. SETUP & DATA LOADING
 # ==========================================
-# Assuming files are in the same root directory as app.py
+# Paths for the final 10k dataset and embeddings
 CSV_PATH = "cleaned_dataset_10k.csv"
 PKL_PATH = "final_embeddings_10k.pkl"
 
 if not os.path.exists(CSV_PATH) or not os.path.exists(PKL_PATH):
-    raise FileNotFoundError("Missing required data files. Ensure CSV and PKL are uploaded.")
+    raise FileNotFoundError("Missing required data files (CSV or PKL) in the root directory.")
 
 # Load the restaurant dataset
 df = pd.read_csv(CSV_PATH)
 df.columns = [c.strip().lower().replace(' ', '_') for c in df.columns]
 
-# Load pre-computed embeddings
+# Load pre-computed embeddings (768 dimensions)
 with open(PKL_PATH, 'rb') as f:
     embedding_data = pickle.load(f)
     dataset_embeddings = embedding_data['embeddings'] if isinstance(embedding_data, dict) else embedding_data
 
-# Load the semantic transformer model (MPNet for high fidelity)
+# Load the semantic transformer model (MPNet)
+# This model converts natural language queries into semantic vectors
 model = SentenceTransformer('sentence-transformers/all-mpnet-base-v2')
 
-# Pre-calculate Persona Taste Profiles (Mean Vectors)
+# Pre-calculate Taste Profiles (Mean Vectors) for each of the 6 personas
 persona_profiles = {}
 for persona in df['reviewer_persona'].unique():
     if pd.isna(persona): continue
@@ -37,14 +38,14 @@ for persona in df['reviewer_persona'].unique():
     persona_profiles[persona] = np.mean(persona_vectors, axis=0)
 
 # ==========================================
-# 2. CORE RECOMMENDATION ENGINE (HYBRID)
+# 2. HYBRID RECOMMENDATION ENGINE
 # ==========================================
 def run_ven_engine(budget, dietary, company, purpose, noise):
     """
-    Finds the best restaurant match using Hybrid Scoring:
-    70% Semantic Contextual Fit + 30% User Rating.
+    Finds the best restaurant match by combining Persona Similarity (General Taste)
+    and Review Similarity (Contextual Fit).
     """
-    # Construct the user's semantic context
+    # Create the user's specific context string
     user_context = f"Searching for a {budget} experience, {dietary} friendly. Group: {company}. Occasion: {purpose}. Atmosphere: {noise}."
     query_vec = model.encode([user_context])
     
@@ -53,25 +54,25 @@ def run_ven_engine(budget, dietary, company, purpose, noise):
                     for p, v in persona_profiles.items()}
     closest_persona = max(persona_sims, key=persona_sims.get)
     
-    # Step B: Filter reviews belonging to that persona
+    # Step B: Search specifically within that persona's reviews for the best contextual fit
     persona_indices = df[df['reviewer_persona'] == closest_persona].index
     persona_embeddings = dataset_embeddings[persona_indices]
     
-    # Step C: Calculate Contextual Similarity for specific reviews within that persona
+    # Compute similarity for every individual review in this persona group
     sub_similarities = cosine_similarity(query_vec, persona_embeddings)[0]
     
     persona_df = df.loc[persona_indices].copy()
     persona_df['semantic_fit'] = sub_similarities
     persona_df['norm_rating'] = persona_df['rating_score'] / 5.0
     
-    # CALCULATE FINAL SCORE
+    # CALCULATE HYBRID SCORE (70% context match + 30% original rating)
     persona_df['final_score'] = (persona_df['semantic_fit'] * 0.7) + (persona_df['norm_rating'] * 0.3)
     
-    # Select the top re-ranked result
+    # Retrieve the top re-ranked result
     top_match = persona_df.sort_values(by='final_score', ascending=False).iloc[0]
     match_pct = int(top_match['final_score'] * 100)
     
-    # Return Styled HTML Result Card
+    # Return Styled HTML Card (with absolute colors to prevent theme overrides)
     return f"""
     <div style="background: white; border-radius: 20px; padding: 25px; color: #0f172a !important; text-align: left; border-left: 10px solid #f97316; box-shadow: 0 10px 25px rgba(0,0,0,0.2);">
         <div style="display:flex; justify-content:space-between; align-items: flex-start;">
@@ -96,57 +97,66 @@ def run_ven_engine(budget, dietary, company, purpose, noise):
     """
 
 # ==========================================
-# 3. GRADIO UI SETUP (FINAL VISIBILITY FIX)
+# 3. GRADIO UI SETUP (VISIBILITY FIX)
 # ==========================================
+# We use targeted CSS to ensure labels are white on dark background, 
+# but choices inside radio buttons are dark on light background.
 ven_css = """
-/* 1. Dark background for the whole app */
+/* Global Container */
 .gradio-container { background-color: #0f172a !important; }
-h1 { color: white !important; text-align: center; font-weight: 900 !important; font-size: 2.5rem !important; margin-bottom: 20px !important; }
+h1 { color: white !important; text-align: center; font-weight: 950 !important; font-size: 2.5rem !important; margin-bottom: 20px !important; }
 
-/* 2. Main Labels (e.g., "3. Social Context") -> MUST BE WHITE */
+/* Input Header Labels (The questions above the inputs) */
 label span { 
     color: white !important; 
     font-weight: 700 !important; 
-    font-size: 15px !important; 
+    font-size: 14px !important; 
+    margin-bottom: 5px !important;
 }
 
-/* 3. RADIO CHOICE TEXT (e.g., "Solo", "Date") -> MUST BE DARK SATE */
-/* Since radio buttons have white backgrounds, white text is invisible. We force it to dark blue/grey. */
-.gr-radio label span, .gr-radio span { 
+/* RADIO CHOICE TEXT (The actual options like 'Solo' or 'Lively') */
+/* These elements are usually inside white/grey boxes in Gradio theme, 
+   so we force the text to be DARK for visibility. */
+.gr-radio label span, 
+.gr-radio span,
+[data-testid="block-info"] + div label span {
     color: #1e293b !important; 
-    font-weight: 600 !important; 
+    font-weight: 700 !important;
 }
 
-/* 4. Orange primary button styling */
+/* Style the primary orange button */
 .ven-button { 
     background-color: #f97316 !important; 
     color: white !important; 
     border: none !important; 
     font-weight: 900 !important; 
     font-size: 18px !important; 
-    height: 50px !important; 
+    height: 52px !important; 
     border-radius: 12px !important;
 }
 
-/* 5. Quick Vibe Starters table colors */
+/* Quick Vibe Starters table styling */
 .gr-samples-table { background-color: #1e293b !important; color: white !important; }
 """
 
-with gr.Blocks(css=ven_css, title="VEN - AI Restaurant Match") as demo:
+with gr.Blocks(css=ven_css, title="VEN - AI Restaurant Matchmaker") as demo:
     gr.Markdown("# 🍔 VEN: Restaurant Matchmaker")
     
     with gr.Row():
         with gr.Column(scale=1):
             with gr.Group():
+                # Survey inputs for the 5 dimensions
                 in_budget = gr.Dropdown(["Budget-friendly", "Mid-range", "Premium"], label="1. Wallet Size", value="Mid-range")
                 in_diet = gr.Dropdown(["Anything", "Vegetarian", "Vegan", "Meat-lover"], label="2. Diet Preference", value="Anything")
                 in_company = gr.Radio(["Solo", "Date/Couple", "Friends", "Business"], label="3. Social Context", value="Date/Couple")
                 in_purpose = gr.Dropdown(["Casual dinner", "Special occasion", "Quick bite", "Professional meeting"], label="4. The Mission", value="Casual dinner")
                 in_noise = gr.Radio(["Quiet/Intimate", "Moderate/Social", "Lively/Music"], label="5. Vibe / Noise", value="Moderate/Social")
+                
                 btn = gr.Button("Find My Table", variant="primary", elem_classes="ven-button")
         
         with gr.Column(scale=1):
-            output_ui = gr.HTML("<div style='text-align:center; padding:100px; color:#64748b; font-weight:700; border: 2px dashed #1e293b; border-radius: 20px;'>Your personalized AI match will appear here</div>")
+            # Placeholder for the AI recommendation card
+            output_ui = gr.HTML("<div style='text-align:center; padding:100px; color:#64748b; font-weight:700; border: 2px dashed #1e293b; border-radius: 20px;'>Fill the survey to generate your AI match</div>")
 
     gr.Markdown("### 🚀 Quick Vibe Starters")
     gr.Examples(
@@ -161,6 +171,7 @@ with gr.Blocks(css=ven_css, title="VEN - AI Restaurant Match") as demo:
         cache_examples=False,
     )
     
+    # Event binding
     btn.click(run_ven_engine, inputs=[in_budget, in_diet, in_company, in_purpose, in_noise], outputs=output_ui)
 
 if __name__ == "__main__":