Search is not available for this dataset
Unnamed: 0
int64
0
7.83M
βŒ€
u
int64
1
3.7M
i
int64
1
3.7M
ts
float64
0
2.24B
label
float64
0
35
idx
int64
1
7.83M
0
384
353
0
0
1
1
411
353
0
0
2
2
676
353
0
0
3
3
467
353
0
0
4
4
120
353
0
0
5
5
149
353
0
0
6
6
204
353
0
0
7
7
53
353
0
0
8
8
342
353
0
0
9
9
80
353
0
0
10
10
523
353
0
0
11
11
412
353
0
0
12
12
158
353
0
0
13
13
536
353
0
0
14
14
711
353
0
0
15
15
712
353
0
0
16
16
6
353
0
0
17
17
81
353
0
0
18
18
457
353
0
0
19
19
719
353
0
0
20
20
381
353
0
0
21
21
505
353
0
0
22
22
356
353
0
0
23
23
218
353
0
0
24
24
643
353
0
0
25
25
217
353
0
0
26
26
610
353
0
0
27
27
705
353
0
0
28
28
410
353
0
0
29
29
58
353
0
0
30
30
566
353
0
0
31
31
708
353
0
0
32
32
103
353
0
0
33
33
456
353
0
0
34
34
231
353
0
0
35
35
468
353
0
0
36
36
282
353
0
0
37
37
734
353
0
0
38
38
527
353
0
0
39
39
363
353
0
0
40
40
603
353
0
0
41
41
174
353
0
0
42
42
469
353
0
0
43
43
692
353
0
0
44
44
366
353
0
0
45
45
401
353
0
0
46
46
662
353
0
0
47
47
539
353
0
0
48
48
518
353
0
0
49
49
15
353
0
0
50
50
650
353
0
0
51
51
682
353
0
0
52
52
139
353
0
0
53
53
592
353
0
0
54
54
183
353
0
0
55
55
343
353
0
0
56
56
66
353
0
0
57
57
136
353
0
0
58
58
380
353
0
0
59
59
167
353
0
0
60
60
195
353
0
0
61
61
597
353
0
0
62
62
392
353
0
0
63
63
398
353
0
0
64
64
428
353
0
0
65
65
451
353
0
0
66
66
111
353
0
0
67
67
630
353
0
0
68
68
114
353
0
0
69
69
242
353
0
0
70
70
345
353
0
0
71
71
413
353
0
0
72
72
402
353
0
0
73
73
593
353
0
0
74
74
513
353
0
0
75
75
497
353
0
0
76
76
702
353
0
0
77
77
8
353
0
0
78
78
672
353
0
0
79
79
198
353
0
0
80
80
334
353
0
0
81
81
663
353
0
0
82
82
563
353
0
0
83
83
601
353
0
0
84
84
254
353
0
0
85
85
714
353
0
0
86
86
596
353
0
0
87
87
344
353
0
0
88
88
572
353
0
0
89
89
675
353
0
0
90
90
112
353
0
0
91
91
143
353
0
0
92
92
185
353
0
0
93
93
361
353
0
0
94
94
65
353
0
0
95
95
573
353
0
0
96
96
332
353
0
0
97
97
197
353
0
0
98
98
326
353
0
0
99
99
32
353
0
0
100

The dataset is dynamic graphs for paper CrossLink. The usage of this dataset can be seen in Github

πŸš€ Introduction

CrossLink learns the evolution pattern of a specific downstream graph and subsequently makes pattern-specific link predictions. It employs a technique called conditioned link generation, which integrates both evolution and structure modeling to perform evolution-specific link prediction. This conditioned link generation is carried out by a transformer-decoder architecture, enabling efficient parallel training and inference. CrossLink is trained on extensive dynamic graphs across diverse domains, encompassing 6 million dynamic edges. Extensive experiments on eight untrained graphs demonstrate that CrossLink achieves state-of-the-art performance in cross-domain link prediction. Compared to advanced baselines under the same settings, CrossLink shows an average improvement of 11.40% in Average Precision across eight graphs. Impressively, it surpasses the fully supervised performance of 8 advanced baselines on 6 untrained graphs.

Architecture

Format

Please keep the dataset in the fellow format:

Unnamed: 0 u i ts label idx
idx-1 source node target node interaction time defalut: 0 from 1 to the #edges

You can prepare those data by the code in preprocess_data folder

You can also use our raw data in huggingface

πŸ“š Citation

If you find this work helpful, please consider citing:

@misc{huang2024graphmodelcrossdomaindynamic,
  title={One Graph Model for Cross-domain Dynamic Link Prediction}, 
  author={Xuanwen Huang and Wei Chow and Yang Wang and Ziwei Chai and Chunping Wang and Lei Chen and Yang Yang},
  year={2024},
  eprint={2402.02168},
  archivePrefix={arXiv},
  primaryClass={cs.LG},
  url={https://arxiv.org/abs/2402.02168}, 
}
Downloads last month
245

Collection including WeiChow/DyGraphs