feat(bulk): SuggestionService.for_selection consensus + POST /api/suggestions/bulk

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

backend/app/services/ml/suggestions.py

@@ -197,3 +197,77 @@ class SuggestionService:
         for cat in result.by_category:
             result.by_category[cat].sort(key=lambda s: s.score, reverse=True)
         return result
+
+    async def for_selection(
+        self,
+        image_ids: list[int],
+        threshold: float = 0.8,
+        top_k: int = 10,
+    ) -> dict[str, list[dict]]:
+        """Consensus suggestions across image_ids. A tag is included iff it
+        was suggested for (or already applied to) >= threshold fraction of
+        the selection AND was acceptable on >= 1 image. Confidence is the
+        mean over images where it was suggested. Aggregated by
+        canonical_tag_id; creates-new (no canonical id) suggestions are
+        skipped (bulk Accept applies by tag id)."""
+        if not image_ids:
+            return {}
+        threshold = min(1.0, max(0.0, threshold))
+        total = len(image_ids)
+
+        stats: dict[int, dict] = {}
+        for image_id in image_ids:
+            sl = await self.for_image(image_id)
+            for category, items in sl.by_category.items():
+                for s in items:
+                    if s.canonical_tag_id is None or s.creates_new_tag:
+                        continue
+                    st = stats.get(s.canonical_tag_id)
+                    if st is None:
+                        st = {
+                            "tag_id": s.canonical_tag_id,
+                            "name": s.display_name,
+                            "category": category,
+                            "source": s.source,
+                            "suggested_count": 0,
+                            "sum_score": 0.0,
+                        }
+                        stats[s.canonical_tag_id] = st
+                    st["suggested_count"] += 1
+                    st["sum_score"] += s.score
+
+        rows = (
+            await self.session.execute(
+                select(
+                    image_tag.c.image_record_id, image_tag.c.tag_id
+                ).where(image_tag.c.image_record_id.in_(image_ids))
+            )
+        ).all()
+        applied_by_tag: dict[int, set[int]] = {}
+        for iid, tid in rows:
+            applied_by_tag.setdefault(tid, set()).add(iid)
+
+        result: dict[str, list[dict]] = {}
+        for st in stats.values():
+            existing_count = len(applied_by_tag.get(st["tag_id"], set()))
+            covered = st["suggested_count"] + existing_count
+            coverage = covered / total
+            if coverage < threshold or st["suggested_count"] < 1:
+                continue
+            result.setdefault(st["category"], []).append(
+                {
+                    "canonical_tag_id": st["tag_id"],
+                    "name": st["name"],
+                    "category": st["category"],
+                    "confidence": round(
+                        st["sum_score"] / st["suggested_count"], 4
+                    ),
+                    "coverage": round(coverage, 4),
+                    "covered_count": covered,
+                    "source": st["source"],
+                }
+            )
+        for cat in result:
+            result[cat].sort(key=lambda x: x["confidence"], reverse=True)
+            result[cat] = result[cat][:top_k]
+        return result