基于物品的协同过滤使用余弦相似度电影推荐系统的Python代码，使用MovieLens数据集

好的，以下是基于物品的协同过滤使用余弦相似度电影推荐系统的Python代码，使用MovieLens数据集：

import pandas as pd
import numpy as np

# 读取数据集
ratings = pd.read_csv('ratings.csv')
movies = pd.read_csv('movies.csv')

# 合并数据集
ratings = pd.merge(ratings, movies, on='movieId')

# 计算电影评分矩阵
movieRatings = ratings.pivot_table(index=['title'], columns=['userId'], values='rating')

# 填充缺失值
movieRatings = movieRatings.fillna(0)

# 计算余弦相似度
from sklearn.metrics.pairwise import cosine_similarity

movieSimilarity = cosine_similarity(movieRatings.T)

# 定义函数，返回与目标电影最相似的K个电影
def topKSimilarMovies(movie, K):
    simMovies = movieSimilarity[movie]
    simMovies = simMovies.argsort()[::-1]
    return simMovies[1:K+1]

# 定义函数，根据最相似的K个电影推荐给用户
def recommendMovies(user, K):
    userRatings = movieRatings[user]
    simMovies = topKSimilarMovies(userRatings.name, K)
    simMoviesRatings = movieRatings.iloc[simMovies]
    weightedRatings = (simMoviesRatings * movieSimilarity[userRatings.name][simMovies]).sum(axis=0)
    weightedRatings = weightedRatings.sort_values(ascending=False)
    moviesToRecommend = weightedRatings.loc[~weightedRatings.index.isin(userRatings.index)]
    return moviesToRecommend.index.tolist()

# 示例：为用户1推荐10部电影
user = 1
K = 10
recommendedMovies = recommendMovies(user, K)
print(recommendedMovies)

注意：上述代码仅作为示例，实际应用时需要根据具体情况进行调整和优化。