-
Notifications
You must be signed in to change notification settings - Fork 3
/
table1_main_results.tex
71 lines (68 loc) · 7.96 KB
/
table1_main_results.tex
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
\begin{table*}[!t]
\caption{Average MRR results for different groups of temporal complex queries. \textbf{X} denotes the variant of TFLEX. \textbf{X(ConE)} replaces the entity part with ConE~\cite{ConE}. \textbf{FLEX} ablates the time part. \textbf{X-1F} merges entity feature and timestamp feature into one feature. \textbf{X-logic} removes the logic part.}
\label{table:main_results}
\centering
\resizebox{\textwidth}{!}{
\begin{tabular}{l l r r r r r r r r r r r r r r r r r r r r r r r r r}
\toprule
\textbf{Dataset} & \textbf{Metrics} & \textbf{Query2box} & \textbf{BetaE} & \textbf{ConE} & \textbf{TFLEX} & \textbf{X(ConE)} & \textbf{FLEX} & \textbf{X-1F} & & \textbf{X-logic} \\ % \textbf{X-entity logic} & \textbf{X-temporal logic}
\midrule
\multirow{8}{*}{ICEWS14} & $\text{avg}_{e} $ & 25.06 & 37.19 & 41.94 & 56.79 & 40.93 & 43.67 & 56.89 & & 56.64 \\ % 57.08 & 56.48
& $\text{avg}_{e , \mathcal{C}_e}$ & & 36.69 & 44.88 & 50.82 & 42.15 & 44.41 & 49.78 & & 51.17 \\ % 51.99 & 51.57
& $\text{avg}_{t} $ & & & & 17.56 & 16.41 & & 18.77 & & 18.03 \\ % 18.10 & 18.20
& $\text{avg}_{t , \mathcal{C}_t}$ & & & & 36.37 & 35.24 & & 37.73 & & 36.39 \\ % 36.36 & 36.63
& $\text{avg}_{\{\mathcal{U}_e\}}$ & & 19.95 & 26.47 & 35.74 & 25.46 & 29.25 & 34.48 & & 34.68 \\ % 35.81 & 35.85
& $\text{avg}_{\{\mathcal{U}_t\}}$ & & & & 26.24 & 24.07 & & 28.04 & & 26.36 \\ % 25.43 & 25.95
& $\text{avg}_{x}$ & & & & 28.03 & 26.65 & & 29.31 & & 28.61 \\ % 28.16 & 29.03
& \textbf{AVG} & & & & 35.93 & 30.13 & & 36.43 & & 35.98 \\ % 36.13 & 36.24
\midrule
\multirow{8}{*}{ICEWS05-15} & $\text{avg}_{e} $ & 24.00 & 31.33 & 40.93 & 48.99 & 36.29 & 38.96 & 49.90 & & 44.80 \\ % 45.34 & 43.25
& $\text{avg}_{e , \mathcal{C}_e}$ & & 29.70 & 43.52 & 46.17 & 38.12 & 42.10 & 46.11 & & 41.92 \\ % 43.62 & 43.20
& $\text{avg}_{t} $ & & & & 4.39 & 4.41 & & 4.43 & & 3.29 \\ % 3.77 & 3.32
& $\text{avg}_{t , \mathcal{C}_t}$ & & & & 30.16 & 29.49 & & 30.26 & & 28.34 \\ % 29.17 & 28.25
& $\text{avg}_{\{\mathcal{U}_e\}}$ & & 21.54 & 43.02 & 54.37 & 36.37 & 44.38 & 54.05 & & 45.36 \\ % 53.21 & 45.88
& $\text{avg}_{\{\mathcal{U}_t\}}$ & & & & 28.69 & 26.40 & & 27.70 & & 23.39 \\ % 23.56 & 24.27
& $\text{avg}_{x}$ & & & & 24.26 & 21.69 & & 24.41 & & 21.95 \\ % 23.33 & 22.63
& \textbf{AVG} & & & & 33.72 & 27.54 & & 33.98 & & 29.86 \\ % 31.71 & 30.11
\midrule
\multirow{8}{*}{GDELT-500} & $\text{avg}_{e} $ & 9.67 & 14.75 & 18.44 & 19.60 & 17.83 & 19.07 & 17.92 & & 17.36 \\ % 17.34 & 17.86
& $\text{avg}_{e , \mathcal{C}_e}$ & & 11.15 & 12.67 & 13.52 & 12.34 & 13.35 & 12.13 & & 12.11 \\ % 12.25 & 12.26
& $\text{avg}_{t} $ & & & & 5.38 & 3.16 & & 5.49 & & 5.75 \\ % 5.14 & 5.79
& $\text{avg}_{t , \mathcal{C}_t}$ & & & & 6.31 & 3.93 & & 6.50 & & 6.86 \\ % 6.02 & 6.97
& $\text{avg}_{\{\mathcal{U}_e\}}$ & & 6.20 & 6.96 & 7.58 & 7.41 & 7.44 & 6.92 & & 6.91 \\ % 6.88 & 7.02
& $\text{avg}_{\{\mathcal{U}_t\}}$ & & & & 6.71 & 6.35 & & 6.59 & & 6.80 \\ % 6.37 & 6.73
& $\text{avg}_{x}$ & & & & 6.17 & 6.17 & & 6.47 & & 6.64 \\ % 6.17 & 6.78
& \textbf{AVG} & & & & 9.32 & 8.17 & & 8.86 & & 8.92 \\ % 8.59 & 9.06
\bottomrule
\end{tabular}
}
\vspace{-4mm}
\end{table*}
% bib
@inproceedings{Query2box,
title={Query2box: Reasoning Over Knowledge Graphs In Vector Space Using Box Embeddings},
author={Ren, H and Hu, W and Leskovec, J},
booktitle={International Conference on Learning Representations (ICLR)},
year={2020}
}
@article{BetaE,
title={Beta Embeddings for Multi-Hop Logical Reasoning in Knowledge Graphs},
author={Ren, Hongyu and Leskovec, Jure},
journal={Neural Information Processing Systems (NeurIPS)},
year={2020}
}
@article{ConE,
title={ConE: Cone Embeddings for Multi-Hop Reasoning over Knowledge Graphs},
author={Zhang, Zhanqiu and Wang, Jie and Chen, Jiajun and Ji, Shuiwang and Wu, Feng},
journal={Advances in Neural Information Processing Systems},
volume={34},
year={2021}
}
% TFLEX
@inproceedings{xueyuan2023tflex,
title={{TFLEX}: Temporal Feature-Logic Embedding Framework for Complex Reasoning over Temporal Knowledge Graph},
author={Lin Xueyuan and Haihong E and Chengjin Xu and Gengxian Zhou and Haoran Luo and Tianyi Hu and Fenglong Su and Ningyuan Li and Mingzhi Sun},
booktitle={Thirty-seventh Conference on Neural Information Processing Systems},
year={2023},
url={https://openreview.net/forum?id=oaGdsgB18L}
}