KG-S2S

Knowledge Is Flat: A Seq2Seq Generative framework For Various Knowledge Graph Completion

Overview of KG-S2S

This repository includes the source code of the paper accepted by COLING'2022.

"Knowledge Is Flat: A Seq2Seq Generative framework For Various Knowledge Graph Completion".

Dependencies

• Compatible with PyTorch 1.11.0 and Python 3.x.
• Dependencies can be installed using requirements.txt.

Dataset:

• We use WN18RR, FB15k-237N, FB15k-237, ICEWS14 and NELL-One dataset for knowledge graph link prediction.
• All the preprocessed data are included in the ./data/processed/ directory. Alternatively, you can download the raw dataset into ./data/raw/ and run ./preprocess.sh to generate the processed data.
• Raw data source:
• WN18RR, FB15k-237
• ICEWS14
• reformatted NELL-One

Pretrained Checkpoint:

To enable a quick evaluation, we upload the trained model. Download the checkpoint folders to ./checkpoint/, and run the evaluation commandline for corresponding dataset.

The results are:

Training and testing:

• Install all the requirements from ./requirements.txt.
• Commands for reproducing the reported results: ##### WN18RR ```shell python3 main.py -dataset 'WN18RR' \ -lr 0.001 \ -epoch 100 \ -batch_size 64 \ -src_descrip_max_length 40 \ -tgt_descrip_max_length 10 \ -use_soft_prompt \ -use_rel_prompt_emb \ -seq_dropout 0.1 \ -num_beams 40 \ -eval_tgt_max_length 30 \ -skip_n_val_epoch 30 \

# evaluation commandline: python3 main.py -dataset 'WN18RR' \ -src_descrip_max_length 40 \ -tgt_descrip_max_length 10 \ -use_soft_prompt \ -use_rel_prompt_emb \ -num_beams 40 \ -eval_tgt_max_length 30 \ -model_path path/to/trained/model \ -use_prefix_search ##### FB15k-237Nshell python3 main.py -dataset 'FB15k-237N' \ -lr 0.001 \ -epoch 50 \ -batch_size 32 \ -src_descrip_max_length 80 \ -tgt_descrip_max_length 80 \ -use_soft_prompt \ -use_rel_prompt_emb \ -seq_dropout 0.2 \ -num_beams 40 \ -eval_tgt_max_length 30 \ -skip_n_val_epoch 30

# evaluation commandline: python3 main.py -dataset 'FB15k-237N' \ -src_descrip_max_length 80 \ -tgt_descrip_max_length 80 \ -use_soft_prompt \ -use_rel_prompt_emb \ -num_beams 40 \ -eval_tgt_max_length 30 \ -model_path path/to/trained/model \ -use_prefix_search
##### FB15k-237shell python3 main.py -dataset 'FB15k-237' \ -lr 0.001 \ -epoch 40 \ -batch_size 32 \ -src_descrip_max_length 80 \ -tgt_descrip_max_length 80 \ -use_soft_prompt \ -use_rel_prompt_emb \ -seq_dropout 0.2 \ -num_beams 40 \ -eval_tgt_max_length 30 \ -skip_n_val_epoch 20

# evaluation commandline: python3 main.py -dataset 'FB15k-237' \ -src_descrip_max_length 80 \ -tgt_descrip_max_length 80 \ -use_soft_prompt \ -use_rel_prompt_emb \ -num_beams 40 \ -eval_tgt_max_length 30 \ -model_path path/to/trained/model \ -use_prefix_search ```

##### ICEWS14 ```shell python3 main.py -dataset 'ICEWS14' \ -lr 0.0005 \ -epoch 100 -batch_size 32 \ -src_descrip_max_length 40 \ -tgt_descrip_max_length 40 \ -temporal \ -use_soft_prompt \ -use_rel_prompt_emb \ -seq_dropout 0.1 \ -num_beams 40 \ -eval_tgt_max_length 26 \ -skip_n_val_epoch 50

# evaluation commandline: python3 main.py -dataset 'ICEWS14' \ -src_descrip_max_length 40 \ -tgt_descrip_max_length 40 \ -temporal \ -use_soft_prompt \ -use_rel_prompt_emb \ -num_beams 40 \ -eval_tgt_max_length 26 \ -model_path path/to/trained/model \ -use_prefix_search
```

##### NELL-One ```shell python3 main.py -dataset 'NELL' \ -lr 0.0005 \ -epoch 30 \ -batch_size 128 \ -src_descrip_max_length 0 \ -tgt_descrip_max_length 0 \ -use_soft_prompt \ -use_rel_prompt_emb \ -num_beams 40 \ -eval_tgt_max_length 25 \ -skip_n_val_epoch 15

# evaluation commandline: python3 main.py -dataset 'NELL' \ -src_descrip_max_length 0 \ -tgt_descrip_max_length 0 \ -use_soft_prompt \ -use_rel_prompt_emb \ -num_beams 40 \ -eval_tgt_max_length 25 \ -model_path path/to/trained/model \ -use_prefix_search
`` --src_descrip_max_lengthdenotes the training batch size --src_descrip_max_lengthdenotes the maximum description length for source entity during training --tgt_descrip_max_lengthdenotes the maximum description length for target entity during training --eval_tgt_max_lengthdenotes the maximum description length for generation during inference --use_soft_promptdenotes the option whether to use soft prompt --use_rel_prompt_embdenotes the option whether to use relation-specific soft prompt (need to enable-use_soft_prompt) --seq_dropoutdenotes the value for sequence dropout --use_prefix_searchdenotes to use constrained decoding method --temporaldenotes the dataset is for temporal knowledge graph completion --skip_n_val_epoch` denotes the number of training epochs without evaluation (evaluation is costly due to the auto-regressive decoding)

Citation

If you used our work or found it helpful, please use the following citation:

@inproceedings{KG_S2S,
    title = "Knowledge Is Flat: A Seq2Seq Generative Framework for Various Knowledge Graph Completion",
    author = "Chen, Chen  and
      Wang, Yufei  and
      Li, Bing  and 
      Lam, Kwok-Yan",
    booktitle = "Proceedings of the 29th International Conference on Computational Linguistics",
}