黑马推荐系统项目实战【三】基于用户的协同过滤 UserCF

用户物品相似度计算

users = ["User1","User2","User3","User4","User5"]
items =["Item A","Item B","Item C","Item D","Item E"]

#用户购买记录数据
datasets = [
    [1, 0, 1, 1, 0],
    [1, 0, 0, 1, 1],
    [1, 0, 1, 0, 0],
    [0, 1, 0, 1, 1],
    [1, 1, 1, 0, 1],
]
import pandas as pd
df = pd.DataFrame(datasets, columns=items,index=users)
df

	Item A	Item B	Item C	Item D	Item E
User1	1	0	1	1	0
User2	1	0	0	1	1
User3	1	0	1	0	0
User4	0	1	0	1	1
User5	1	1	1	0	1

from sklearn.metrics.pairwise import pairwise_distances
# 计算用户的相似度
user_similar = 1 - pairwise_distances(df.values, metric='jaccard')
user_similar = pd.DataFrame(user_similar, columns=users, index=users)
user_similar

D:\Apps\Anaconda3\lib\site-packages\sklearn\metrics\pairwise.py:1735: DataConversionWarning: Data was converted to boolean for metric jaccard
warnings.warn(msg, DataConversionWarning)

	User1	User2	User3	User4	User5
User1	1.000000	0.50	0.666667	0.2	0.4
User2	0.500000	1.00	0.250000	0.5	0.4
User3	0.666667	0.25	1.000000	0.0	0.5
User4	0.200000	0.50	0.000000	1.0	0.4
User5	0.400000	0.40	0.500000	0.4	1.0

# 计算物品的相似度
item_similar = 1 - pairwise_distances(df.T.values, metric='jaccard')
item_similar = pd.DataFrame(item_similar, columns=items, index=items)
item_similar

D:\Apps\Anaconda3\lib\site-packages\sklearn\metrics\pairwise.py:1735:DataConversionWarning: Data was converted to boolean for metric jaccard
warnings.warn(msg, DataConversionWarning)

	Item A	Item B	Item C	Item D	Item E
Item A	1.00	0.200000	0.75	0.40	0.400000
Item B	0.20	1.000000	0.25	0.25	0.666667
Item C	0.75	0.250000	1.00	0.20	0.200000
Item D	0.40	0.250000	0.20	1.00	0.500000
Item E	0.40	0.666667	0.20	0.50	1.000000

基于用户的协同过滤 UserCF

# 为每一个用户找到最相似的2个用户
topN_users = {}
for i in user_similar.index:
# 取出每一列数据 删除自己 按照相似度排序
    _df = user_similar.loc[i].drop([i])
    _df_sorted = _df.sort_values(ascending = False)
    top2 = list(_df_sorted.index[:2])
    topN_users[i] = top2

topN_users

{‘User1’: [‘User3’, ‘User2’],
‘User2’: [‘User4’, ‘User1’],
‘User3’: [‘User1’, ‘User5’],
‘User4’: [‘User2’, ‘User5’],
‘User5’: [‘User3’, ‘User4’]}

topN_users.items()

dict_items([(‘User1’, [‘User3’, ‘User2’]), (‘User2’, [‘User4’, ‘User1’]), (‘User3’, [‘User1’, ‘User5’]), (‘User4’, [‘User2’, ‘User5’]), (‘User5’, [‘User3’, ‘User4’])])

df.loc['User1'].index

Index([‘Item A’, ‘Item B’, ‘Item C’, ‘Item D’, ‘Item E’], dtype=‘object’)

import numpy as np
# 根据topn的相似用户构建推荐结果
re_results = {}
for user, sim_users in topN_users.items():
    re_result = set() # 当前用户的相似用户的交互过的物品集合
    for sim_user in sim_users:
        re_result = re_result.union(set(df.loc[sim_user].replace(0, np.nan).dropna().index))
        
    # 过滤掉自己交互过的物品
    re_result -= set(df.loc[user].replace(0,np.nan).dropna().index)
    re_results[user] = re_result

re_results

{‘User1’: {‘Item E’},
‘User2’: {‘Item B’, ‘Item C’},
‘User3’: {‘Item B’, ‘Item D’, ‘Item E’},
‘User4’: {‘Item A’, ‘Item C’},
‘User5’: {‘Item D’}}