Sequential_Recommendation_Tensorflow/make_datasets.py at master · slientGe/Sequential_Recommendation_Tensorflow · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
import pandas as pd


def make_datasets(data, len_Seq, len_Tag, len_Pred):

    #file_path = 'input/u.data'

    p = data.groupby('item')['user'].count().reset_index().rename(columns={'user':'item_count'})
    data = pd.merge(data,p,how='left',on='item')
    data = data[data['item_count'] > 5].drop(['item_count'],axis=1)

    # ReMap item ids
    item_unique = data['item'].unique().tolist()
    item_map = dict(zip(item_unique, range(1,len(item_unique) + 1)))
    item_map[-1] = 0
    all_item_count = len(item_map)
    data['item'] = data['item'].apply(lambda x: item_map[x])

    # ReMap usr ids
    user_unique = data['user'].unique().tolist()
    user_map = dict(zip(user_unique, range(1, len(user_unique) + 1)))
    user_map[-1] = 0
    all_user_count = len(item_map)
    data['user'] = data['user'].apply(lambda x: user_map[x])

    # Get user session
    data = data.sort_values(by=['user','timestamps']).reset_index(drop=True)

    # 生成用户序列
    user_sessions = data.groupby('user')['item'].apply(lambda x: x.tolist()) \
        .reset_index().rename(columns={'item': 'item_list'})

    train_users = []
    train_seqs = []
    train_targets = []

    test_users = []
    test_seqs = []
    test_targets = []

    items_usr_clicked = {}

    for index, row in user_sessions.iterrows():
        user = row['user']
        items = row['item_list']


        test_item = items[-1*len_Pred :]
        test_seq = items[-1* (len_Pred + len_Seq) :-1*len_Pred]
        test_users.append(user)
        test_seqs.append(test_seq)
        test_targets.append(test_item)


        train_build_items = items[:-1*len_Pred]

        items_usr_clicked[user] = train_build_items

        for i in range(len_Seq, len(train_build_items) - len_Tag + 1):
            item = train_build_items[i:i+ len_Tag]
            seq = train_build_items[max(0,i - len_Seq):i]

            train_users.append(user)
            train_seqs.append(seq)
            train_targets.append(item)

    d_train = pd.DataFrame({'user':train_users,'seq':train_seqs,'target':train_targets})

    d_test = pd.DataFrame({'user': test_users, 'seq': test_seqs, 'target': test_targets})

    d_info= (all_user_count, all_item_count, items_usr_clicked, user_map, item_map)

    return d_train,d_test,d_info


if __name__ == '__main__':
    make_datasets(5,3,2)


'''

src data

196	242	3	881250949
186	302	3	891717742
22	377	1	878887116
244	51	2	880606923
166	346	1	886397596
298	474	4	884182806
115	265	2	881171488
253	465	5	891628467
305	451	3	886324817
6	86	3	883603013
62	257	2	879372434
286	1014	5	879781125
200	222	5	876042340
210	40	3	891035994
224	29	3	888104457


train_data
                          seq            target  user
0     [1, 290, 492, 381, 752]    [467, 523, 11]     1
1   [290, 492, 381, 752, 467]    [523, 11, 673]     1
2   [492, 381, 752, 467, 523]   [11, 673, 1046]     1
3    [381, 752, 467, 523, 11]  [673, 1046, 650]     1
4    [752, 467, 523, 11, 673]  [1046, 650, 378]     1
5   [467, 523, 11, 673, 1046]   [650, 378, 180]     1
6   [523, 11, 673, 1046, 650]   [378, 180, 390]     1
7   [11, 673, 1046, 650, 378]   [180, 390, 666]     1
8  [673, 1046, 650, 378, 180]   [390, 666, 513]     1
9  [1046, 650, 378, 180, 390]   [666, 513, 432]     1

test_data
                            seq       target  user
0    [633, 657, 1007, 948, 364]  [522, 0, 0]     1
1       [26, 247, 49, 531, 146]   [32, 0, 0]     2
2      [459, 477, 369, 770, 15]  [306, 0, 0]     3
3  [1093, 946, 1101, 690, 1211]  [526, 0, 0]     4
4     [732, 266, 669, 188, 253]  [986, 0, 0]     5
5      [410, 446, 104, 782, 96]   [26, 0, 0]     6
6     [146, 817, 536, 694, 186]  [525, 0, 0]     7
7      [395, 669, 281, 289, 98]  [731, 0, 0]     8
8     [588, 671, 369, 292, 304]  [250, 0, 0]     9
9        [472, 222, 82, 716, 8]  [131, 0, 0]    10


'''