双12智能批量计算推荐

import numpy as np
from sklearn.metrics.pairwise import cosine_similarity

# 假设我们有一个用户-商品评分矩阵
ratings = np.array([
    [5, 3, 0, 1],
    [4, 0, 0, 1],
    [1, 1, 0, 5],
    [1, 0, 0, 4]
])

# 计算用户之间的相似度
user_similarity = cosine_similarity(ratings)

# 推荐函数
def recommend(user_id, ratings, similarity, k=2):
    user_ratings = ratings[user_id]
    similar_users = similarity[user_id].argsort()[::-1][1:k+1]
    recommendations = np.zeros(ratings.shape[1])
    
    for similar_user in similar_users:
        recommendations += similarity[user_id][similar_user] * ratings[similar_user]
    
    recommendations[user_ratings.nonzero()] = 0  # 排除已评分的商品
    return recommendations.argsort()[::-1]

# 示例调用
recommended_items = recommend(0, ratings, user_similarity)
print("Recommended items for user 0:", recommended_items)