Ensembling Graph Predictions for AMR Parsing

Authors: Thanh Lam Hoang, Gabriele Picco, Yufang Hou, Young-Suk Lee, Lam Nguyen, Dzung Phan, Vanessa Lopez, Ramon Fernandez Astudillo

NeurIPS 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental To validate our approach, we carried out experiments in AMR parsing problems. The experimental results demonstrate that the proposed approach can combine the strength of state-of-the-art AMR parsers to create new predictions that are more accurate than any individual models in five standard benchmark datasets.
Researcher Affiliation Industry 1 IBM Research, Dublin, Ireland 2 IBM Research, Thomas J. Watson Research Center, Yorktown Heights, USA
Pseudocode Yes Algorithm 1: Graph ensemble with the Graphene algorithm.
Open Source Code Yes Source code is open-sourced1. 1https://github.com/IBM/graph_ensemble_learning
Open Datasets Yes Similarly to [Bevilacqua et al., 2021], we use five standard benchmark datasets [dat] to evaluate our approach. Table 1 shows the statistics of the datasets. AMR 2.0 and AMR 3.0 are divided into train, development and testing sets and we use them for in-distribution evaluation in Section 4.2. (...) AMR benchmark datasets. https://amr.isi.edu/download.html.
Dataset Splits Yes Table 1: Benchmark datasets. (...) For AMR 2.0 and 3.0, the models are trained on the training dataset, validated on the development dataset. We report results on testing sets in the in-distribution evaluation.
Hardware Specification Yes In all experiments, we used a Tesla GPU V100 for model training and used 8 CPUs for making an ensemble.
Software Dependencies No The paper mentions software components and models like BART, T5, ADAM optimization, and Stanford Core NLP, but does not provide specific version numbers for these dependencies.
Experiment Setup Yes The model is trained with 30 epochs. We use ADAM optimization with a learning rate of 1e-4 and a mini-batch size of 4.