基于物品的协同过滤Base Item CF (使用jaccard_score 距离评价标准)

import  pandas as pd
import  pprint
import  numpy as np

users = ['user1','user2','user3','user4','user5',]
items = ['item A','item B','item C','item D','item E',]


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]
]


df = pd.DataFrame(datasets,columns= items,index= users)

print(df)

from sklearn.metrics import  jaccard_similarity_score,jaccard_score

# sim = jaccard_similarity_score(df['item A'],df['item B'])
sim = jaccard_score(df['item A'],df['item B'])
print(sim)


from  sklearn.metrics.pairwise import  pairwise_distances

user_similar = 1 - pairwise_distances(df.values,metric='jaccard')
# print(user_similar)

user_similar = pd.DataFrame(user_similar,columns=users,index=users)
print(user_similar)


items_similar = 1 - pairwise_distances(df.T.values,metric='jaccard')

items_similar = pd.DataFrame(items_similar,columns=items,index=items)
print(items_similar)


topN_items = {}
# print(user_similar.index)
for i in items_similar.index:
    _df = items_similar.loc[i].drop([i])
    print(_df)
    _df_sorted = _df.sort_values(ascending=False)
    top2 = list(_df_sorted.index[:2])
    topN_items[i] = top2

print('Top 2 相似物品:')
print(topN_items)


rs_results = {}

for user in df.index:
    rs_result = set()
    for item in df.loc[user].replace(0,np.nan).dropna().index: #取出每个用户当前已经购买物品列表
         #根据每个物品 找出最相似的TOP_N物品，构建初始推荐结果
         rs_result = rs_result.union(topN_items[item])
    #过滤掉用户自己已经购买的物品
    rs_result -= set(df.loc[user].replace(0,np.nan).dropna().index)
    rs_results[user] = rs_result
print('最终推荐结果:')
print(rs_results)

最终推荐结果:
{'user1': {'item B', 'item E'}, 'user2': {'item C', 'item B'}, 'user3': {'item B', 'item E'}, 'user4': {'item A'}, 'user5': {'item D'}}