Frame Semantic Role Labeling Using Arbitrary-Order Conditional Random Fields
Authors: Chaoyi Ai, Kewei Tu
AAAI 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Our experiments show that our approach achieves state-of-the-art performance in FSRL. |
| Researcher Affiliation | Academia | Chaoyi Ai, Kewei Tu* School of Information Science and Technology, Shanghai Tech University Shanghai Engineering Research Center of Intelligent Vision and Imaging {aichy,tukw}@shanghaitech.edu.cn |
| Pseudocode | No | The paper describes the method using prose and equations, but no structured pseudocode or algorithm blocks are present. |
| Open Source Code | Yes | 1Code: https://github.com/aichy98/Frame SRL-AAAI24 |
| Open Datasets | Yes | We used the benchmark datasets Frame Net versions 1.5 and 1.72, hereafter referred to as FN1.5 and FN1.7, respectively, to evaluate the effectiveness of our models. FN1.5 is widely used in previous research. FN1.7 is more comprehensive than FN1.5 and is known for its extended semantic content. We adhered to the train/dev/test split used in prior work (Peng et al. 2018). We utilized Frame Net s exemplar sentences, annotations linked to frames and their lexical units as supplemental training data, a practice frequently adopted in preceding researches (Chen, Zheng, and Chang 2021; Bastianelli, Vanzo, and Lemon 2020; Zheng et al. 2022; Zheng, Wang, and Chang 2023). |
| Dataset Splits | Yes | We adhered to the train/dev/test split used in prior work (Peng et al. 2018). |
| Hardware Specification | No | The paper does not specify any particular hardware components (e.g., CPU, GPU models, memory, or cloud instance types) used for running the experiments. |
| Software Dependencies | No | The paper mentions using 'bert-base-uncased' as the pretrained language model but does not list specific software dependencies with their version numbers (e.g., Python, PyTorch, TensorFlow versions). |
| Experiment Setup | No | We do grid search for hyperparameter tuning and details of our hyper-parameter settings can be found in the supplementary material. |