سیستم توصیه گر فیلم

میخام این پیاده سازی انجام بشه . User-Item Interaction (MovieLens) │ ▼ Preprocessing & Mapping - Implicit/Explicit feedback - Map MovieLens IDs → TMDb metadata │ ▼ Base Recommender (RecBole) ┌─────────────────────────────┐ │ LightGCN │ ← Collaborative Filtering │ SASRec │ ← Sequential Recommendation └─────────────────────────────┘ │ ▼ Top-K Recommendations per User - Top 10 (configurable) - Scores from model predictions │ ▼ LLM (GPT, LLaMA, etc.) - Input: Metadata (title, genre, overview) - Input: User interaction history - Output: Human-readable explanations │ ▼ Final Recommendation + Explanation - For each user: list of items + generated explanations