MARLPaR

Code and models for the paper Path Reasoning over Knowledge Graph: A Multi-Agent and Reinforcement Learning Based Method

We development the code based on the code of MINERVA [Go for a Walk and Arrive at the Answer - Reasoning over Paths in Knowledge Bases using Reinforcement Learning] (https://github.com/shehzaadzd/MINERVA)

MARLPaR is a multi-agent system which answers queries in a knowledge graph of entities and relations. Starting from an entity node, MARLPaR has two agents to carry out relation selection and entity selection, respectively, in an iterative manner.

Requirements

To install the various python dependences (including tensorflow)

pip install -r requirements.txt

Training

The hyperparam configs for each experiments are in the configs directory. To start a particular experiment, just do

sh run.sh configs/${dataset}.sh

where the ${dataset}.sh is the name of the config file. For example,

sh run.sh configs/WN18RR.sh

Testing

make

load_model=1
model_load_dir="saved_models/WN18RR/model.ckpt"

Citation

If you use this code, please cite our paper

@inproceedings{li2018path,
  title={Path reasoning over knowledge graph: A multi-agent and reinforcement learning based method},
  author={Li, Zixuan and Jin, Xiaolong and Guan, Saiping and Wang, Yuanzhuo and Cheng, Xueqi},
  booktitle={2018 IEEE International Conference on Data Mining Workshops (ICDMW)},
  pages={929--936},
  year={2018},
  organization={IEEE}
}