M-Walk: Learning to Walk over Graphs using Monte Carlo Tree Search

Authors: Yelong Shen, Jianshu Chen, Po-Sen Huang, Yuqing Guo, Jianfeng Gao

NeurIPS 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results on several graph-walking benchmarks show that M-Walk is able to learn better policies than other RL-based methods, which are mainly based on policy gradients. M-Walk also outperforms traditional KBC baselines.
Researcher Affiliation Industry 1Tencent AI Lab, Bellevue, WA, USA. {yelongshen, jianshuchen}@tencent.com 2Microsoft Research, Redmond, WA, USA {yuqguo, jfgao}@microsoft.com
Pseudocode No The paper describes the training and prediction algorithms in text but does not include any formally labeled 'Pseudocode' or 'Algorithm' blocks.
Open Source Code Yes The code of this paper is available at: https://github.com/yelongshen/Graph Walk
Open Datasets Yes We use WN18RR and NELL995 knowledge graph datasets for evaluation. WN18RR [6] is created from the original WN18 [2]... The NELL995 dataset was released by [38]...
Dataset Splits Yes We use the same data split and preprocessing protocol as in [6] for WN18RR and in [38, 5] for NELL995.
Hardware Specification No The paper mentions 'Cuda' and 'Tensor Flowgpu' which implies GPU usage for computations, but it does not specify any particular GPU models, CPU models, or other detailed hardware specifications for the experimental setup.
Software Dependencies No The paper mentions software like 'Tensor Flowgpu' and 'Cuda' and implementation languages like 'C++', but it does not provide specific version numbers for any software dependencies or libraries.
Experiment Setup No The paper states: 'We briefly describe the tasks here, and give the experiment details and hyperparameters in Appendix B.' and 'The details of the hyperparameters for M-Walk are described in Appendix B.2.2 of the supplementary material.' This indicates that the specific experimental setup details, including hyperparameters, are deferred to the supplementary material and are not present in the main text.