P9-Code

Information

Models

• DE-TransE: transformation model
• DE-SimplE: tensor decompositional model
• DE-DistMult: tensor decompositional model
• ATiSE: tensor decompositional model
• TERO: transformation model
• TFLEX: neural network model

Metrics

• MRR
• HITS@1
• HITS@3
• HITS@10

Data Prepare

In this project, we use JSON as the file to transfer the data.

The script is in ~/github/P9-Code/dataprepare/corruptedquadruple/corrupted_quadruple_generate.py

Usage on Terminal

1. Move the directory to ~/github/P9-Code/dataprepare/corruptedquadruple
2. Run the script

python corrupted_quadruple_generate.py

Usage on CLAAUDIA

1. Login CLAAUDIA

ssh -l <aau email> ai-fe02.srv.aau.dk

2. Transfer files into CLAAUDIA

scp some-file <aau email>@ai-fe02.srv.aau.dk:~

3. Run Singualrity

srun --pty singularity shell --fakeroot --writable <image name>

4. Run the script

python3 <script name>

5. Download files from CLAAUDIA (if it is a directory, add -r after scp)

scp <aau email>@ai-fe02.srv.aau.dk:~/some-folder/some-subfolder/some-file .

Dataset Source

ATISE

• ICEWS14
• $YAGO11K_{D}$

TERO

• ICEWS14
• ICEWS05-15
• YAGO11K
• Wikidata12k

Diachronic Embedding

• ICEWS14
• ICEWS05-15
• GDELT