Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Query Your Model with Definitions in FrameNet: An Effective Method for Frame Semantic Role Labeling
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Add FE definitions as extra training data (ablation) | Aged · FN 1.7 test | F1-score: 74.73 -> 75.37 | Helped |
| Add FE definitions as extra training data (ablation) | Aged · FN 1.5 test | F1-score: 74.28 -> 74.84 | Helped |
| Remove label markers (ablation) | Aged · FN 1.5 test, no exemplar | F1-score: 74.28 -> 74.22 | No clear effect |
| Add target markers around the target word (ablation) | Aged · FN 1.5 test, no exemplar | F1-score: 63.74 -> 74.28 | Helped |
| Use FrameNet definitions instead of simple role-specific questions | Aged · FN 1.5 test, no exemplar | F1: 73.69 -> 74.28 | Helped |
| Use frame definitions rather than one FE definition per query | Aged · FN 1.5 test, no exemplar | F1: 74.23 -> 74.28 | No clear effect |
| Jointly train on frame identification and FSRL | Aged · FN 1.5 test, no exemplar | FSRL F1-score: 74.28 -> 74.46 (+0.18) | Helped |
| Train Aged with no instances of the evaluated frames | Aged · FN 1.5, zero-shot frames | F1-score: 82.45 -> 86.26 | Hurt |
| Train on FN 1.5 and evaluate on FN 1.7 rather than train and evaluate on FN 1.7 | Aged · FN 1.7 test | F1-score: 74.73 -> 73.72 | Hurt |
| Use frame definitions to query all arguments in one pair rather than FE-specific pairs | Aged · FN 1.5 test, no exemplar | number of text-definition pairs: 44252 -> 4458 | Helped |
Closest papers by meaning
- Frame Semantic Role Labeling Using Arbitrary-Order Conditional Random Fields (AAAI 2024 · similarity 0.88)
- Dependency or Span, End-to-End Uniform Semantic Role Labeling (AAAI 2019 · similarity 0.88)
- Span-Based Semantic Role Labeling with Argument Pruning and Second-Order Inference (AAAI 2022 · similarity 0.88)
- MAPLE: Many-Shot Adaptive Pseudo-Labeling for In-Context Learning (ICML 2025 · similarity 0.81)
- Logic-in-Frames: Dynamic Keyframe Search via Visual Semantic-Logical Verification for Long Video Understanding (NeurIPS 2025 · similarity 0.81)
- SRTube: Video-Language Pre-Training with Action-Centric Video Tube Features and Semantic Role Labeling (CVPR 2024 · similarity 0.81)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.