推荐系统(4)--生成电影推荐

生成电影推荐

代码如下:

# -*- coding:utf-8 -*-
'''
   生成电影推荐系统
'''
import json
import numpy as np

# 定义一个计算两个用户之间欧几里得距离的函数 user1和user2
def euclidean_score(dataset, user1, user2):
    if user1 not in dataset:
        raise TypeError('User'+user1+'not present in the dataset')
    if user2 not in dataset:
        raise TypeError('User'+user2+'not present in the dataset')
    # 提取两个用户均评过分的电影
    rated_by_both = {}
    for item in dataset[user1]:
        if item in dataset [user2]:
            rated_by_both[item]=1
    # 如果两个用户均没评分过,则得分为0
    if len(rated_by_both)==0:
        return 0
    # 计算平方和的平方根 并将其归一化 使得评分取值在0-1之间
    squared_difference = []
    for item in dataset[user1]:
        if item in dataset[user2]:
            squared_difference.append(np.square(dataset[user1][item]-dataset[user2][item]))
        return 1/(1+np.sqrt(np.sum(squared_difference)))

# 皮尔逊系数函数
def pearson_score(dataset,user1,user2):
    if user1 not in dataset:
        raise TypeError('User'+user1+'not present in the dataset')
    if user2 not in dataset:
        raise TypeError('User'+user2+'not present in the dataset')
    # 提取两个用户都评过分的电影
    rated_by_both = {}
    for item in dataset[user1]:
       if item in dataset[user2]:
           rated_by_both[item]=1
    num_ratings=len(rated_by_both)
    # 如果两个用户都没有评分 则说明两个用户之间没有相似度 返回值为0
    if num_ratings==0:
        return  0
    # 计算相同评分电影的平方值之和
    user1_sum=np.sum([dataset[user1][item] for item in rated_by_both])
    user2_sum=np.sum([dataset[user2][item] for item in rated_by_both])

    # 计算相同电影的评分的平方和
    user1_squared_sum=np.sum([np.square(dataset[user1][item]) for item in rated_by_both])
    user2_squared_sum=np.sum([np.square(dataset[user2][item]) for item in rated_by_both])

    # 计算数据集的乘积之和
    product_sum=np.sum([dataset[user1][item]*dataset[user2][item] for item in rated_by_both])

    # 计算皮尔逊相关系数
    Sxy=product_sum-(user1_sum*user2_sum/num_ratings)
    Sxx=user1_squared_sum-np.square(user1_sum)/num_ratings
    Syy=user2_squared_sum-np.square(user2_sum)/num_ratings

    # 考虑分母为0的情况
    if Sxx*Syy==0:
        return 0
    # 如果上述正常 返回皮尔逊相关系数
    return Sxy/np.sqrt(Sxx*Syy)

# 寻找特定数量的与输入用户相似的用户
# dataset:代表数据库 user: 输入用户 num_users:相似的用户个数
def find_similar_users(dataset,user,num_users):
    if user not in dataset:
        raise TypeError('User '+user+' not present in the dataset')

    # 计算所有用户的皮尔逊相关度
    scores = np.array([[x,pearson_score(dataset,user,x)] for x in dataset if user!=x])

    # 评分按照第二列进行排列

    scores_sorted =np.argsort(scores[:,1])

    #  评分按照降序排列
    scores_sorted_dec=scores_sorted[::-1]

    #  提取多个高分并返回
    top_k=scores_sorted_dec[0:num_users]
    return scores[top_k]

# 定义一个为用户生成电影推荐的函数 首先检查用户是否在数据库中
def generate_recommendations(dataset, user):
    if user not in dataset:
        raise TypeError('User ' + user + ' not present in the dataset')

    total_scores = {}
    similarity_sums = {}

    for u in [x for x in dataset if x != user]:
        similarity_score = pearson_score(dataset, user, u)

        if similarity_score <= 0:
            continue

        for item in [x for x in dataset[u] if x not in dataset[user] or dataset[user][x] == 0]:
            total_scores.update({item: dataset[u][item] * similarity_score})
            similarity_sums.update({item: similarity_score})

    if len(total_scores) == 0:
        return ['No recommendations possible']

    # Create the normalized list
    movie_ranks = np.array([[total/similarity_sums[item], item]
            for item, total in total_scores.items()])

    # Sort in decreasing order based on the first column
    movie_ranks = movie_ranks[np.argsort(movie_ranks[:, 0])[::-1]]

    # Extract the recommended movies
    recommendations = [movie for _, movie in movie_ranks]

    return recommendations
if __name__=='__main__':
    data_file='movie_ratings.json'
    with open(data_file,'r') as f:
        data=json.loads(f.read())
    # 为Michael Henry 生成推荐
    user='Michael Henry'
    print "\nRecommendations for "+user+":"
    movies=generate_recommendations(data,user)
    for i,movie in enumerate(movies):
        print str(i+1)+'. '+movie
    # 用户John Carson 看过所有电影 因此生成0推荐
    user='John Carson'
    print "\nRecommendation for "+user+":"
    movies=generate_recommendations(data,user)
    for i,movie in enumerate(movies):
        print str(# -*- coding:utf-8 -*-
'''
   生成电影推荐系统
'''
import json
import numpy as np

# 定义一个计算两个用户之间欧几里得距离的函数 user1和user2
def euclidean_score(dataset, user1, user2):
    if user1 not in dataset:
        raise TypeError('User'+user1+'not present in the dataset')
    if user2 not in dataset:
        raise TypeError('User'+user2+'not present in the dataset')
    # 提取两个用户均评过分的电影
    rated_by_both = {}
    for item in dataset[user1]:
        if item in dataset [user2]:
            rated_by_both[item]=1
    # 如果两个用户均没评分过,则得分为0
    if len(rated_by_both)==0:
        return 0
    # 计算平方和的平方根 并将其归一化 使得评分取值在0-1之间
    squared_difference = []
    for item in dataset[user1]:
        if item in dataset[user2]:
            squared_difference.append(np.square(dataset[user1][item]-dataset[user2][item]))
        return 1/(1+np.sqrt(np.sum(squared_difference)))

# 皮尔逊系数函数
def pearson_score(dataset,user1,user2):
    if user1 not in dataset:
        raise TypeError('User'+user1+'not present in the dataset')
    if user2 not in dataset:
        raise TypeError('User'+user2+'not present in the dataset')
    # 提取两个用户都评过分的电影
    rated_by_both = {}
    for item in dataset[user1]:
       if item in dataset[user2]:
           rated_by_both[item]=1
    num_ratings=len(rated_by_both)
    # 如果两个用户都没有评分 则说明两个用户之间没有相似度 返回值为0
    if num_ratings==0:
        return  0
    # 计算相同评分电影的平方值之和
    user1_sum=np.sum([dataset[user1][item] for item in rated_by_both])
    user2_sum=np.sum([dataset[user2][item] for item in rated_by_both])

    # 计算相同电影的评分的平方和
    user1_squared_sum=np.sum([np.square(dataset[user1][item]) for item in rated_by_both])
    user2_squared_sum=np.sum([np.square(dataset[user2][item]) for item in rated_by_both])

    # 计算数据集的乘积之和
    product_sum=np.sum([dataset[user1][item]*dataset[user2][item] for item in rated_by_both])

    # 计算皮尔逊相关系数
    Sxy=product_sum-(user1_sum*user2_sum/num_ratings)
    Sxx=user1_squared_sum-np.square(user1_sum)/num_ratings
    Syy=user2_squared_sum-np.square(user2_sum)/num_ratings

    # 考虑分母为0的情况
    if Sxx*Syy==0:
        return 0
    # 如果上述正常 返回皮尔逊相关系数
    return Sxy/np.sqrt(Sxx*Syy)

# 寻找特定数量的与输入用户相似的用户
# dataset:代表数据库 user: 输入用户 num_users:相似的用户个数
def find_similar_users(dataset,user,num_users):
    if user not in dataset:
        raise TypeError('User '+user+' not present in the dataset')

    # 计算所有用户的皮尔逊相关度
    scores = np.array([[x,pearson_score(dataset,user,x)] for x in dataset if user!=x])

    # 评分按照第二列进行排列

    scores_sorted =np.argsort(scores[:,1])

    #  评分按照降序排列
    scores_sorted_dec=scores_sorted[::-1]

    #  提取多个高分并返回
    top_k=scores_sorted_dec[0:num_users]
    return scores[top_k]

# 定义一个为用户生成电影推荐的函数 首先检查用户是否在数据库中
def generate_recommendations(dataset, user):
    if user not in dataset:
        raise TypeError('User ' + user + ' not present in the dataset')

    total_scores = {}
    similarity_sums = {}
    # 计算用户与数据库中其他用户的皮尔逊相关系数
    for u in [x for x in dataset if x != user]:
        similarity_score = pearson_score(dataset, user, u)

        if similarity_score <= 0:
            continue
    # 找到还未被用户评分过的电影
        for item in [x for x in dataset[u] if x not in dataset[user] or dataset[user][x] == 0]:
            total_scores.update({item: dataset[u][item] * similarity_score})
            similarity_sums.update({item: similarity_score})
    # 如果用户看过所有电影 就不能为其推荐电影
    if len(total_scores) == 0:
        return ['No recommendations possible']

    # 生成一个电影评分标准化列表
    movie_ranks = np.array([[total/similarity_sums[item], item]
            for item, total in total_scores.items()])

    # 根据第一列对皮尔逊系数进行降序排列
    movie_ranks = movie_ranks[np.argsort(movie_ranks[:, 0])[::-1]]

    # 提取出推荐的电影
    recommendations = [movie for _, movie in movie_ranks]

    return recommendations
if __name__=='__main__':
    data_file='movie_ratings.json'
    with open(data_file,'r') as f:
        data=json.loads(f.read())
    # 为Michael Henry 生成推荐
    user='Michael Henry'
    print "\nRecommendations for "+user+":"
    movies=generate_recommendations(data,user)
    for i,movie in enumerate(movies):
        print str(i+1)+'. '+movie
    # 用户John Carson 看过所有电影 因此生成0推荐
    user='John Carson'
    print "\nRecommendation for "+user+":"
    movies=generate_recommendations(data,user)
    for i,movie in enumerate(movies):
        print str(i+1)+'. '+movie

输出结果如下:

Recommendations for Michael Henry:
1. Jerry Maguire
2. Anger Management
3. Inception

Recommendation for John Carson:
1. No recommendations possible

使用json数据如下:

{
    "John Carson": 
    {
        "Inception": 2.5,
        "Pulp Fiction": 3.5,
        "Anger Management": 3.0,
        "Fracture": 3.5,
        "Serendipity": 2.5,
        "Jerry Maguire": 3.0
    },
    "Michelle Peterson": 
    {
        "Inception": 3.0,
        "Pulp Fiction": 3.5,
        "Anger Management": 1.5,
        "Fracture": 5.0,
        "Jerry Maguire": 3.0,
        "Serendipity": 3.5 
    },
    "William Reynolds": 
    {
        "Inception": 2.5,
        "Pulp Fiction": 3.0,
        "Fracture": 3.5,
        "Jerry Maguire": 4.0
    },
    "Jillian Hobart": 
    {
        "Pulp Fiction": 3.5,
        "Anger Management": 3.0,
        "Jerry Maguire": 4.5,
        "Fracture": 4.0,
        "Serendipity": 2.5 
    },
    "Melissa Jones": 
    {
        "Inception": 3.0,
        "Pulp Fiction": 4.0,
        "Anger Management": 2.0,
        "Fracture": 3.0,
        "Jerry Maguire": 3.0,
        "Serendipity": 2.0
    },
    "Alex Roberts": 
    {
       "Inception": 3.0,
       "Pulp Fiction": 4.0,
       "Jerry Maguire": 3.0,
       "Fracture": 5.0,
       "Serendipity": 3.5
    },
    "Michael Henry": 
    {
       "Pulp Fiction": 4.5,
       "Serendipity": 1.0,
       "Fracture": 4.0
    }
}