refactor: update strong signal item percentage and simplify scoring parameters

Author: TimilsinaBimal

app/services/profile/sampling.py

@@ -69,8 +69,8 @@ def sample_items(
             if not (it.get("_is_loved") or it.get("_is_liked") or it.get("_id") in added_item_ids)
         ]
 
-        # Always include strong signal items: Loved/Liked: 30%, Added: 20%
-        strong_signal_items = loved_liked_items[: int(max_items * 0.45)] + added_items[: int(max_items * 0.20)]
+        # Always include strong signal items: Loved/Liked: 40%, Added: 20%
+        strong_signal_items = loved_liked_items[: int(max_items * 0.40)] + added_items[: int(max_items * 0.20)]
         strong_signal_scored = [self.scoring_service.process_item(it) for it in strong_signal_items]
 
         # Score watched items and sort by score

app/services/recommendation/all_based.py

@@ -119,8 +119,6 @@ async def get_recommendations_from_all_items(
                         profile=profile,
                         scorer=self.scorer,
                         mtype=mtype,
-                        is_ranked=False,
-                        is_fresh=False,
                     )
 
                     # Apply genre multiplier (if whitelist available)

app/services/recommendation/scoring.py

@@ -92,20 +92,15 @@ def m_raw(year: int | None) -> float:
         return m_raw, alpha
 
     @staticmethod
-    def apply_quality_adjustments(
-        score: float, wr: float, vote_count: int, popularity: float, is_ranked: bool, is_fresh: bool
-    ) -> float:
+    def apply_quality_adjustments(score: float, wr: float, vote_count: int, popularity: float) -> float:
         """Apply simple quality boost for high-confidence items only."""
-        # Simplified: only boost proven high-quality items, no penalties
-        # Trust the weighted rating formula to handle low vote counts naturally
         if vote_count >= 1000 and wr >= 7.5 and popularity <= MAXIMUM_POPULARITY_SCORE:
             # Proven gem: high confidence, high quality
             return score * 1.10
         elif vote_count >= 500 and wr >= 7.0 and popularity <= MAXIMUM_POPULARITY_SCORE:
             # Good confidence and quality
             return score * 1.05
 
-        # Everything else: trust the base scoring
         return score
 
     @staticmethod
@@ -114,8 +109,6 @@ def calculate_final_score(
         profile: Any,
         scorer: Any,
         mtype: str,
-        is_ranked: bool = False,
-        is_fresh: bool = False,
     ) -> float:  # noqa: E501
         """
         Calculate final recommendation score combining profile similarity and quality.
@@ -125,8 +118,6 @@ def calculate_final_score(
             profile: User taste profile
             scorer: ProfileScorer instance
             mtype: Media type (movie/tv) to determine minimum rating
-            is_ranked: Whether item is from ranked source
-            is_fresh: Whether item should get freshness boost
             minimum_rating_tv: Minimum rating constant for TV
             minimum_rating_movie: Minimum rating constant for movies
 
@@ -155,8 +146,6 @@ def calculate_final_score(
         # light boost for high-confidence items (no penalties!)
         vote_count = item.get("vote_count", 0)
         popularity = item.get("popularity", 0)
-        final_score = RecommendationScoring.apply_quality_adjustments(
-            base_score, wr, vote_count, popularity, is_ranked=is_ranked, is_fresh=is_fresh
-        )
+        final_score = RecommendationScoring.apply_quality_adjustments(base_score, wr, vote_count, popularity)
 
         return final_score

app/services/recommendation/theme_based.py

@@ -111,8 +111,6 @@ async def get_recommendations_for_theme(
                         profile=profile,
                         scorer=self.scorer,
                         mtype=mtype,
-                        is_ranked=False,
-                        is_fresh=False,
                     )
 
                     # Apply genre multiplier (if whitelist available)

app/services/recommendation/top_picks.py

@@ -1,4 +1,5 @@
 import asyncio
+import time
 from collections import defaultdict
 from datetime import date, datetime, timedelta
 from typing import Any
@@ -17,6 +18,7 @@
 )
 from app.services.profile.sampling import SmartSampler
 from app.services.profile.scorer import ProfileScorer
+from app.services.recommendation.filtering import RecommendationFiltering
 from app.services.recommendation.metadata import RecommendationMetadata
 from app.services.recommendation.scoring import RecommendationScoring
 from app.services.recommendation.utils import content_type_to_mtype, filter_watched_by_imdb, resolve_tmdb_id
@@ -70,7 +72,6 @@ async def get_top_picks(
         Returns:
             List of recommended items
         """
-        import time
 
         start_time = time.time()
 
@@ -98,15 +99,11 @@ async def get_top_picks(
         scored_candidates = []
         for item in filtered_candidates:
             try:
-                is_ranked = item.get("_ranked_candidate", False)
-                is_fresh = item.get("_fresh_boost", False)
                 final_score = RecommendationScoring.calculate_final_score(
                     item=item,
                     profile=profile,
                     scorer=self.scorer,
                     mtype=mtype,
-                    is_ranked=is_ranked,
-                    is_fresh=is_fresh,
                 )
                 scored_candidates.append((final_score, item))
             except Exception as e:
@@ -214,8 +211,6 @@ async def _fetch_discover_with_profile(
         Returns:
             List of candidate items
         """
-        # Get excluded genres
-        from app.services.recommendation.filtering import RecommendationFiltering
 
         excluded_genre_ids = RecommendationFiltering.get_excluded_genre_ids(self.user_settings, content_type)
         without_genres = "|".join(str(g) for g in excluded_genre_ids) if excluded_genre_ids else None
@@ -371,7 +366,7 @@ def _apply_diversity_caps(
         Apply diversity caps to ensure balanced results.
 
         Caps:
-        - Recency: max 15% from items released in last 6 months
+        - Recency: max 15% from items released in last year
         - Genre: max 50% per genre
         - Creator: max 3 items per creator
         - Era: max 50% per era