EMNLP Findings2023
Coarse-to-Fine Dual Encoders are Better Frame Identification Learners
Kaikai An, Ce Zheng, Bofei Gao, Haozhe Zhao, Baobao Chang
Abstract
Frame identification aims to find semantic frames associated with target words in a sentence. Recent researches measure the similarity or matching score between targets and candidate frames by modeling frame definitions. However, they either lack sufficient representation learning of the definitions or face challenges in efficiently selecting the most suitable frame from over 1000 candidate frames. Moreover, commonly used lexicon filtering (lf ) to obtain candidate frames for the target may ignore out-of-vocabulary targets and cause inadequate frame modeling. In this paper, we propose COFFTEA, a Coarse-to-Fine Frame and Target Encoders Architecture. With contrastive learning and dual encoders, COFFTEA efficiently and effectively models the alignment between frames and targets. By employing a coarse-to-fine curriculum learning procedure, COFFTEA gradually learns to differentiate frames with varying degrees of similarity. Experimental results demonstrate that COFFTEA outperforms previous models by 0.93 overall scores and 1.53 R@1 without lf. Further analysis suggests that COFFTEA can better model the relationships between frame and frame, as well as target and target. The code for our approach is available at https://github.com/ pkunlp-icler/COFFTEA.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
3 top-tier papers have cited it since. Know what they found before your reviewers do.
Lune can draw on
the full paper · the top-tier papers that cite it · related work
Free to start. No credit card required.
Lineage
- Can LLMs Extract Frame-Semantic Arguments? cites Joint Multi-Decoder Framework with Hierarchical Pointer Network for Frame Semantic Parsing
- Can LLMs Extract Frame-Semantic Arguments? cites A Knowledge-Guided Framework for Frame Identification
- Can LLMs Extract Frame-Semantic Arguments? cites Query Your Model with Definitions in FrameNet: An Effective Method for Frame Semantic Role Labeling
- Query Your Model with Definitions in FrameNet: An Effective Method for Frame Semantic Role Labeling cites Joint Multi-Decoder Framework with Hierarchical Pointer Network for Frame Semantic Parsing
- Query Your Model with Definitions in FrameNet: An Effective Method for Frame Semantic Role Labeling cites A Knowledge-Guided Framework for Frame Identification
Cited by top-tier papers3
- MMICL: Empowering Vision-language Model with Multi-Modal In-Context LearningHaozhe Zhao, Zefan Cai, Shuzheng Si, Xiaojian Ma et al.ICLR 2024 · 206 citations
- Hierarchical Retrieval: The Geometry and a Pretrain-Finetune RecipeChong You, Rajesh Jayaram, Ananda Theertha Suresh, Robin Nittka et al.NeurIPS 2025 · 4 citations
- Can LLMs Extract Frame-Semantic Arguments?Jacob Daniel Devasier, Rishabh Mediratta, Chengkai LiEMNLP 2025 · 1 citation
Builds on6
- Supervised Contrastive LearningPrannay Khosla, Piotr Teterwak, Chen Wang, Aaron Sarna et al.NeurIPS 2020 · 7,049 citations
- SimCSE: Simple Contrastive Learning of Sentence EmbeddingsTianyu Gao, Xingcheng Yao, Danqi ChenEMNLP 2021 · 2,496 citations
- Joint Multi-Decoder Framework with Hierarchical Pointer Network for Frame Semantic ParsingXudong Chen, Ce Zheng, Baobao ChangACL Findings 2021 · 11 citations
- Query Your Model with Definitions in FrameNet: An Effective Method for Frame Semantic Role LabelingCe Zheng, Yiming Wang, Baobao ChangAAAI 2023 · 6 citations
- A Knowledge-Guided Framework for Frame IdentificationXuefeng Su, Ru Li, Xiaoli Li, Jeff Z. Pan et al.ACL 2021
Related papers
- A Graph-Based Neural Model for End-to-End Frame Semantic ParsingZhichao Lin, Yueheng Sun, Meishan ZhangEMNLP 2021 · 9 citations
- Robust Frame-Semantic Models with Lexical Unit Trees and Negative SamplesJacob Daniel Devasier, Yogesh Gurjar, Chengkai LiACL 2024
- Verb Sense Clustering using Contextualized Word Representations for Semantic Frame InductionKosuke Yamada, Ryohei Sasano, Koichi TakedaACL Findings 2021 · 4 citations
- Intent Discovery with Frame-guided Semantic Regularization and AugmentationYajing Sun, Rui Zhang, Jingyuan Yang, Wei PengACL Findings 2023
- CATE: A Contrastive Pre-trained Model for Metaphor Detection with Semi-supervised LearningZhenxi Lin, Qianli Ma, Jiangyue Yan, Jieyu ChenEMNLP 2021 · 15 citations
