Create app.py

app.py

@@ -0,0 +1,124 @@
+import gradio as gr
+import pandas as pd
+import numpy as np
+import os
+from sentence_transformers import SentenceTransformer
+from sklearn.metrics.pairwise import cosine_similarity
+
+# --- Configuration & Model ---
+model = SentenceTransformer('all-MiniLM-L6-v2')
+MAIN_DATASET = 'bitewise_clean_dataset.csv'
+DISH_EMBEDDINGS = 'BiteWise_Dish_Embeddings.npy'
+USER_REVIEWS_DB = 'BiteWise_User_Reviews.csv'
+
+df_main = pd.read_csv(MAIN_DATASET)
+dish_embeddings = np.load(DISH_EMBEDDINGS)
+
+# --- Custom CSS for Styling ---
+custom_css = """
+#main-container {background-color: #fcfcfc;}
+.gradio-container {max-width: 900px !important; margin: auto;}
+.primary-btn {
+    background: linear-gradient(90deg, #ff5f6d 0%, #ffc371 100%) !important;
+    border: none !important;
+    color: white !important;
+    font-weight: bold !important;
+    border-radius: 50px !important;
+}
+.primary-btn:hover {transform: scale(1.02); transition: 0.3s;}
+.dish-card {
+    background: white; 
+    border-radius: 15px; 
+    padding: 20px; 
+    margin-bottom: 15px; 
+    box-shadow: 0 4px 15px rgba(0,0,0,0.05);
+    border-left: 6px solid #ff5f6d;
+}
+.status-msg {
+    padding: 10px;
+    border-radius: 8px;
+    font-weight: 500;
+}
+"""
+
+def execute_search(query):
+    current_df = pd.concat([df_main, pd.read_csv(USER_REVIEWS_DB)] if os.path.exists(USER_REVIEWS_DB) else [df_main], ignore_index=True)
+    query_vec = model.encode([query])
+    sim_scores = cosine_similarity(query_vec, dish_embeddings).flatten()
+    current_df['score'] = sim_scores
+    results = current_df.sort_values(by='score', ascending=False).head(3)
+    
+    html = "<div style='margin-top: 20px;'>"
+    for _, row in results.iterrows():
+        confidence = int(min(99, 85 + (row['score'] * 15)))
+        html += f"""
+        <div class="dish-card">
+            <div style="display: flex; justify-content: space-between; align-items: center;">
+                <h3 style="margin: 0; color: #2D3436;">🍴 {row['dish_name']}</h3>
+                <span style="background: #FFF0F0; color: #FF5F6D; padding: 4px 12px; border-radius: 20px; font-size: 0.8em; font-weight: bold;">{confidence}% Match</span>
+            </div>
+            <p style="margin: 5px 0; color: #636E72;">📍 <b>{row['restaurant_name']}</b></p>
+            <p style="margin: 10px 0; font-style: italic; color: #2D3436; line-height: 1.4;">"{row['taste_review']}"</p>
+            <div style="display: flex; gap: 10px; font-size: 0.85em; color: #B2BEC3;">
+                <span>⭐ {row['rating']}</span> | <span>📷 {row.get('visual_description', 'No photo desc')}</span>
+            </div>
+        </div>
+        """
+    html += "</div>"
+    return html
+
+def save_and_reset(dish, rest, rate, taste, visual, user):
+    if not dish or not rest:
+        return "⚠️ Please fill in the dish and restaurant name.", dish, rest, rate, taste, visual, user
+    
+    new_entry = {'dish_name': dish, 'restaurant_name': rest, 'rating': rate, 'taste_review': taste, 'visual_description': visual, 'user_name': user}
+    new_df = pd.DataFrame([new_entry])
+    
+    if os.path.exists(USER_REVIEWS_DB):
+        existing = pd.read_csv(USER_REVIEWS_DB)
+        pd.concat([existing, new_df], ignore_index=True).to_csv(USER_REVIEWS_DB, index=False)
+    else:
+        new_df.to_csv(USER_REVIEWS_DB, index=False)
+    
+    global dish_embeddings
+    dish_embeddings = np.vstack([dish_embeddings, model.encode([f"{dish} {taste}"])])
+    
+    return f"✨ <b>Amazing!</b> {user}, '{dish}' has been added to our flavor map.", "", "", 4.5, "", "", ""
+
+# --- UI Layout ---
+with gr.Blocks(css=custom_css, theme=gr.themes.Default(primary_hue="orange", secondary_hue="gray")) as demo:
+    with gr.Column(elem_id="main-container"):
+        gr.HTML("""
+            <div style="text-align: center; padding: 20px 0;">
+                <h1 style="font-size: 2.5em; color: #2D3436; margin-bottom: 0;">BiteWise <span style="color: #ff5f6d;">AI</span></h1>
+                <p style="color: #636E72; font-size: 1.1em;">The Intelligence Behind Your Next Craving</p>
+            </div>
+        """)
+        
+        with gr.Tabs():
+            with gr.Tab("🔍 DISCOVER"):
+                query_input = gr.Textbox(label="What are you craving?", placeholder="e.g., A rich, creamy pasta with lots of cheese", lines=1)
+                search_btn = gr.Button("Find My Flavor", variant="primary", elem_classes="primary-btn")
+                gr.Examples(examples=["Crispy chicken burger with spicy aioli", "Fresh summer salad with mango", "Deep dish pizza with pepperoni"], inputs=query_input)
+                output_area = gr.HTML()
+                search_btn.click(execute_search, query_input, output_area)
+
+            with gr.Tab("✨ CONTRIBUTE"):
+                gr.Markdown("### Share your latest culinary find")
+                with gr.Row():
+                    user_in = gr.Textbox(label="Your Name", placeholder="Chef...")
+                    dish_in = gr.Textbox(label="Dish Name", placeholder="The star of the show")
+                    rest_in = gr.Textbox(label="Restaurant", placeholder="Where was it?")
+                with gr.Row():
+                    rate_in = gr.Slider(1, 5, value=4.5, step=0.1, label="Rating")
+                    taste_in = gr.Textbox(label="Taste Notes", placeholder="Describe the flavors...")
+                    visual_in = gr.Textbox(label="Visual Plating", placeholder="How did it look?")
+                
+                save_btn = gr.Button("Sync Review", variant="primary", elem_classes="primary-btn")
+                status_out = gr.HTML(elem_classes="status-msg")
+                
+                save_btn.click(save_and_reset, 
+                               inputs=[dish_in, rest_in, rate_in, taste_in, visual_in, user_in], 
+                               outputs=[status_out, dish_in, rest_in, rate_in, taste_in, visual_in, user_in])
+
+demo.launch()