Polish merchant spotlight and category inputs

app.py

@@ -322,6 +322,16 @@ def _tags_for_category(category: str) -> List[str]:
     return sorted({tag for p in products for tag in p.get("tags", [])})
 
 
+def _category_to_metadata_text(category: str) -> str:
+    category = str(category or "").strip()
+    if not category:
+        return "Clothing"
+    for product in PRODUCTS:
+        if product.get("category") == category and product.get("categories"):
+            return product["categories"]
+    return f"Clothing > {category}"
+
+
 def update_product_choices(category: str):
     names = _product_names_for_category(category)
     return gr.update(choices=names, value=names[0] if names else None)
@@ -397,7 +407,7 @@ def _predict_product(product_name: str) -> Dict[str, Any]:
 
 
 def _predict_custom(review: str, features: str, categories: str, price: Any, rating: Any, count: Any) -> Dict[str, Any]:
-    meta = {"features_text": features or "", "categories_text": categories or "", "price": _safe_float(price), "average_rating": _safe_float(rating), "rating_number": _safe_float(count)}
+    meta = {"features_text": features or "", "categories_text": _category_to_metadata_text(categories), "price": _safe_float(price), "average_rating": _safe_float(rating), "rating_number": _safe_float(count)}
     return _predictor().predict(review or "No review text provided.", meta)
 
 
@@ -749,6 +759,38 @@ def export_merchant_scores(rows):
 
 
 def negative_product_spotlight(limit: int = 6) -> str:
+    cards = []
+    for aspect in ASPECTS:
+        best = None
+        for product in PRODUCTS:
+            try:
+                result = _predict_product(product["name"])
+            except Exception:
+                continue
+            d = result.get("aspect_details", {}).get(aspect, {})
+            if d.get("label") != "Negative":
+                continue
+            conf = _safe_float(d.get("confidence"))
+            hits = _keyword_hits(product.get("review", ""), aspect)
+            reason = ", ".join(hits[:3]) if hits else _short_text(product.get("review") or product.get("features"), 48)
+            if best is None or conf > best[0]:
+                best = (conf, product, reason)
+        if best is None:
+            cards.append(
+                f'<div class="negative-card compact-negative"><h4>{aspect}</h4>'
+                '<div class="empty-negative">No strong negative sample</div></div>'
+            )
+            continue
+        conf, product, reason = best
+        img = f'<img class="mini-product-img" src="{_esc(_product_image_url(product))}" alt="{_esc(product["name"])}">'
+        cards.append(
+            f'<div class="negative-card compact-negative"><h4>{aspect}</h4>'
+            f'<div class="negative-body">{img}<div><b>{_esc(_short_text(product["name"], 42))}</b>'
+            f'<div class="small-label">{_esc(product["category"])} | conf {conf:.2f}</div>'
+            f'<p>{_esc(_short_text(reason, 56))}</p></div></div></div>'
+        )
+    return '<div class="negative-grid">' + "".join(cards[:limit]) + '</div>'
+
     items = []
     for product in PRODUCTS:
         try:
@@ -865,7 +907,7 @@ def _read_uploaded_records(file_obj) -> List[Dict[str, Any]]:
 def download_new_product_template():
     rows = [[
         "Cotton Polyester Blend, Slim Fit, Graphic Print, Machine Wash",
-        "Clothing > Women > Tops > T-Shirts",
+        "T-Shirts",
         29.99,
         4.1,
         35,
@@ -873,7 +915,7 @@ def download_new_product_template():
     ]]
     return _write_csv_download(
         "new_product_metadata_template.csv",
-        ["features", "categories", "price", "average_rating", "rating_number", "focus_aspect"],
+        ["features", "category", "price", "average_rating", "rating_number", "focus_aspect"],
         rows,
     )
 
@@ -885,7 +927,7 @@ def batch_screen_new_products(file_obj) -> Tuple[str, List[List[Any]]]:
     rows = []
     for i, data in enumerate(records, 1):
         features = data.get("features") or data.get("features_text") or ""
-        categories = data.get("categories") or data.get("categories_text") or ""
+        categories = data.get("category") or data.get("categories") or data.get("categories_text") or ""
         price = _safe_float(data.get("price"), 0.0)
         rating = _safe_float(data.get("average_rating") or data.get("rating"), 4.0)
         count = _safe_float(data.get("rating_number") or data.get("rating_count"), 0.0)
@@ -945,7 +987,7 @@ def download_external_review_template():
     rows = [[
         "The fabric is soft and the color looks good, but it runs small.",
         "Cotton Blend, Slim Fit, Zipper Closure",
-        "Clothing > Women > Jackets",
+        "Jackets",
         39.99,
         4.2,
         312,
@@ -953,7 +995,7 @@ def download_external_review_template():
     ]]
     return _write_csv_download(
         "external_review_template.csv",
-        ["review", "features", "categories", "price", "average_rating", "rating_number", "highlight_aspect"],
+        ["review", "features", "category", "price", "average_rating", "rating_number", "highlight_aspect"],
         rows,
     )
 
@@ -966,7 +1008,7 @@ def batch_external_review_predict(file_obj) -> Tuple[str, str, List[List[Any]]]:
     for i, data in enumerate(records, 1):
         review = data.get("review") or data.get("review_text") or ""
         features = data.get("features") or data.get("features_text") or ""
-        categories = data.get("categories") or data.get("categories_text") or ""
+        categories = data.get("category") or data.get("categories") or data.get("categories_text") or ""
         price = _safe_float(data.get("price"), 0.0)
         rating = _safe_float(data.get("average_rating") or data.get("rating"), 4.0)
         count = _safe_float(data.get("rating_number") or data.get("rating_count"), 0.0)
@@ -1060,7 +1102,14 @@ mark { background:#fde68a; color:#111827; border-radius:4px; padding:1px 3px; }
 .negative-grid { display:grid; grid-template-columns:repeat(6, minmax(0, 1fr)); gap:10px; margin:8px 0 14px; }
 .negative-card { display:grid; grid-template-columns:64px 1fr; gap:10px; border:1px solid #fecaca; border-left:4px solid #ef4444; border-radius:8px; padding:10px; background:#fffafa; }
 .negative-card p { margin:5px 0 0; color:#334155; font-size:13px; }
+.compact-negative { display:block; min-height:178px; }
+.compact-negative h4 { margin:0 0 8px; color:#b91c1c; font-size:14px; letter-spacing:.02em; }
+.negative-body { display:grid; grid-template-columns:54px 1fr; gap:8px; align-items:start; }
+.negative-body b { display:block; font-size:13px; line-height:1.25; }
+.negative-body p { font-size:12px; line-height:1.35; }
+.empty-negative { color:#64748b; font-size:12px; border:1px dashed #fecaca; border-radius:8px; padding:14px 8px; background:white; }
 .mini-product-img { width:64px; height:76px; object-fit:cover; border-radius:6px; border:1px solid var(--line); background:#f8fafc; }
+.negative-body .mini-product-img { width:54px; height:64px; }
 @media (max-width:1100px) { .negative-grid { grid-template-columns:repeat(3, minmax(0, 1fr)); } }
 @media (max-width:860px) { .aspect-grid, .metric-grid, .product-layout, .decision-layout, .negative-grid { grid-template-columns:1fr; } .product-img { width:100%; height:180px; } }
 """
@@ -1068,6 +1117,8 @@ mark { background:#fde68a; color:#111827; border-radius:4px; padding:1px 3px; }
 
 def build_app() -> gr.Blocks:
     categories = ["All"] + sorted({p["category"] for p in PRODUCTS})
+    merchant_categories = categories[1:] or ["Clothing"]
+    default_merchant_category = merchant_categories[0]
     tags = sorted({tag for p in PRODUCTS for tag in p.get("tags", [])})
     with gr.Blocks(css=CSS, title="Clothing Sentiment Analysis") as demo:
         gr.HTML('<div id="hero"><div class="eyebrow">BERT + Metadata Cross-Attention</div><h1>Clothing Review Sentiment Intelligence App</h1><p>Explore overall sentiment, six aspect-level opinions, metadata-driven risks, and updated 10W experiment reports in a compact customer decision-support prototype.</p><div class="hero-pills"><span>Consumer decision support</span><span>Merchant diagnostics</span><span>Research dashboard</span></div></div>')
@@ -1107,7 +1158,6 @@ def build_app() -> gr.Blocks:
                 gr.HTML('<div class="module-head"><h3>Product Score Monitor</h3><span class="info-tip">i<span class="tip-content">' + model_info_html() + '</span></span></div>')
                 gr.HTML('<div class="small-label">Most negative products across the six aspects</div>')
                 negative_spotlight = gr.HTML('<div class="note-card">Loading most negative products...</div>')
-                aspect_overview = gr.Dataframe(headers=["Aspect", "Negative Count", "Most Negative Product", "Confidence", "Reason"], value=merchant_aspect_overview(), datatype=["str", "number", "str", "number", "str"], interactive=False, label="Six-aspect negative overview")
                 with gr.Accordion("Score filters and export", open=True):
                     with gr.Row():
                         merchant_metric = gr.Dropdown(["Overall"] + ASPECTS, value="Overall", label="Aspect")
@@ -1124,7 +1174,7 @@ def build_app() -> gr.Blocks:
                         with gr.Column(scale=1):
                             with gr.Group(visible=True) as risk_single_group:
                                 new_features = gr.Textbox("Cotton Polyester Blend, Slim Fit, Graphic Print, Machine Wash", label="New product features", lines=4)
-                                new_categories = gr.Textbox("Clothing > Women > Tops > T-Shirts", label="New product categories")
+                                new_categories = gr.Dropdown(merchant_categories, value=default_merchant_category, label="New product category")
                                 with gr.Row():
                                     new_price = gr.Number(29.99, label="Price")
                                     new_rating = gr.Number(4.1, label="Expected or early average rating")
@@ -1151,7 +1201,7 @@ def build_app() -> gr.Blocks:
                                 external_review = gr.Textbox("The fabric is soft and the color looks good, but it runs small and the zipper feels weak.", label="External customer review", lines=5)
                                 with gr.Row():
                                     ext_features = gr.Textbox("Cotton Blend, Slim Fit, Zipper Closure", label="Product features", lines=3)
-                                    ext_categories = gr.Textbox("Clothing > Women > Jackets", label="Product categories", lines=3)
+                                    ext_categories = gr.Dropdown(merchant_categories, value=default_merchant_category, label="Product category")
                                 with gr.Row():
                                     ext_price = gr.Number(39.99, label="Price")
                                     ext_rating = gr.Number(4.2, label="Average rating")