Update calculate_cosine_similarity.py

calculate_cosine_similarity.py

@@ -1,12 +1,36 @@
-from sklearn.metrics.pairwise import cosine_similarity
+from pymongo import MongoClient
 import numpy as np
+from sklearn.metrics.pairwise import cosine_similarity
+
+client = MongoClient("mongodb+srv://waseoke:[email protected]/test?retryWrites=true&w=majority")
+db = client["two_tower_model"]
+user_embedding_collection = db["user_embeddings"]
+train_dataset = db["train_dataset"]
+
+def calculate_similarity(user_id):
+    # 사용자 임베딩 가져오기
+    user_data = user_embedding_collection.find_one({"user_id": user_id})
+    if not user_data:
+        raise ValueError(f"No embedding found for user_id: {user_id}")
+    
+    user_embedding = np.array(user_data["embedding"]).reshape(1, -1)
+
+    # Anchor 데이터 가져오기
+    anchor_embeddings = []
+    train_data = list(train_dataset.find())
+    for entry in train_data:
+        anchor_embeddings.append(entry["anchor_embedding"])
+    
+    anchor_embeddings = np.array(anchor_embeddings)
 
-def calculate_cosine_similarity(user_embedding, product_embeddings, product_ids, top_n=5):
-    user_embedding = user_embedding.reshape(1, -1)
-    product_embeddings = np.array(product_embeddings)
+    # Cosine Similarity 계산
+    similarities = cosine_similarity(user_embedding, anchor_embeddings).flatten()
 
-    similarities = cosine_similarity(user_embedding, product_embeddings).flatten()
-    top_indices = similarities.argsort()[::-1][:top_n]
-    recommendations = [(product_ids[i], similarities[i]) for i in top_indices]
+    # 가장 유사한 anchor 선택
+    most_similar_index = np.argmax(similarities)
+    most_similar_entry = train_data[most_similar_index]
+    most_similar_positive = most_similar_entry["positive"]
 
-    return recommendations
+    print(f"Most similar anchor for user {user_id}: {most_similar_entry['anchor']}")
+    print(f"Recommended positive product: {most_similar_positive}")
+    return most_similar_positive