Encoder-Decoder Based Unified Semantic Role Labeling with Label-Aware Syntax

Authors: Hao Fei, Fei Li, Bobo Li, Donghong Ji12794-12802

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

Reproducibility Variable Result LLM Response
Research Type Experimental Empirical experiments show that our framework significantly outperforms all existing graph-based methods on the Co NLL09 and Universal Proposition Bank datasets.
Researcher Affiliation Academia Hao Fei, Fei Li, Bobo Li, Donghong Ji* Key Laboratory of Aerospace Information Security and Trusted Computing, Ministry of Education, School of Cyber Science and Engineering, Wuhan University, Wuhan, China {hao.fei, boboli, dhji}@whu.edu.cn, foxlf823@gmail.com
Pseudocode No The paper does not contain structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide concrete access to source code for the methodology described in this paper. It only links to the BERT repository, which is a third-party tool.
Open Datasets Yes We train and evaluate all models on two SRL benchmarks, including Co NLL09 (English), and Universal Proposition Bank (eight languages).
Dataset Splits Yes We employ the official training, development and test sets in each dataset.
Hardware Specification No The paper does not provide specific hardware details (e.g., GPU/CPU models, memory amounts) used for running its experiments.
Software Dependencies No The paper mentions using BERT (base-cased version) and an Adam optimizer, but does not provide specific software dependencies like programming language or library versions (e.g., Python 3.x, PyTorch 1.x).
Experiment Setup Yes In terms of hyper-parameters, since BERT is used, the size of word representations is 768. The size of POS tag embeddings is 50. We use a 3-layer LA-GCN with 350 hidden units, and the output size of LSTM decoder is 300. We adopt the Adam optimizer with an initial learning rate 2e-5, mini-batch size 16 and regularization weight 0.12.