This repository contains the code for the paper "CapsKG: Enabling Continual Knowledge Integration in Language Models for Automatic Knowledge Graph Completion" https://link.springer.com/chapter/10.1007/978-3-031-47240-4_33
If you use this code in your research, please cite the following paper:
Omeliyanenko, J., Zehe, A., Hotho, A., Schlör, D. (2023). CapsKG: Enabling Continual Knowledge Integration in Language Models for Automatic Knowledge Graph Completion. In: Payne, T.R., et al. The Semantic Web – ISWC 2023. ISWC 2023. Lecture Notes in Computer Science, vol 14265. Springer, Cham. https://doi.org/10.1007/978-3-031-47240-4_33
Experiments were run using python 3.8
Pip dependencies are listed in the requirements.txt file.
The data used in the experiments are contained in the dat folder.
Used prompts are contained within the specific dataset folders in templates.json.
The code can be run using the run.py script.
Run arguments may be adjusted in the config.py file.
The following arguments are used for the specific models run in the paper:
- BERT:
--backbone bert --baseline one_mlm - Adapter:
--backbone bert_adapter --baseline one_mlm - BERT-CL:
--backbone bert --baseline ncl_mlm - Adapter-CL:
--backbone bert_adapter --baseline ncl_mlm - CapsKG:
--backbone bert_adapter --baseline ctr_kg
Available Datasets are:
- WN18:
--task wn18 - YAGO3-10:
--task yago - FB15k:
--task fbl
Example usage:
python3 run.py --bert_model bert-base-uncased --baseline ctr_kg --backbone bert_adapter --task wn18 --scenario "til_classification" --use_predefine_args --ntasks 5 --num_train_epochs 1 --yaml_param_num 0 --output_dir "\my_output_directory" --auto_resume --reuse_data