Research map / Label-efficient and robust learning / Domain adaptation and generalization
Evaluation of Semantic Dependency Labeling Across Domains
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 Communicator domain configuration (UT and LA tables) to the generic model (ablation) | OnePar generic · Communicator target labeling | F-measure: 0.57 -> 0.68 | Helped |
| add Communicator domain configuration (UT and LA tables) to the generic model (ablation) | OnePar generic · Communicator lenient argument labeling | F-measure: 0.25 -> 0.30 | Helped |
| add Communicator domain configuration to the generic model (ablation) | OnePar generic · Communicator lenient argument identification | recall: 0.44 -> 0.34 | Hurt |
| add about 300 Communicator sentences (10%) to FrameNet training data (ablation) | OnePar with domain configuration · Communicator target labeling | F-measure: 0.68 -> 0.88 | Helped |
| add about 300 Communicator sentences (10%) to FrameNet training data (ablation) | OnePar with domain configuration · Communicator lenient argument labeling | F-measure: 0.30 -> 0.58 | Helped |
| train only on 10% Communicator data instead of FrameNet plus 10% Communicator (ablation) | OnePar · Communicator lenient argument labeling | F-measure: 0.58 -> 0.63 | Helped |
| train only on 10% Communicator data instead of FrameNet plus 10% Communicator (ablation) | OnePar · Communicator target labeling | F-measure: 0.88 -> 0.89 | No clear effect |
| increase Communicator-only training data from 10% to 50% (ablation) | OnePar domain-specific · Communicator lenient argument labeling | F-measure: 0.63 -> 0.78 | Helped |
| increase Communicator-only training data from 10% to 50% (ablation) | OnePar domain-specific · Communicator target labeling | F-measure: 0.89 -> 0.92 | Helped |
| move generic FrameNet-trained model from FrameNet test data to Communicator dialogs (ablation) | OnePar generic · target labeling | F-measure: 0.64 -> 0.57 | Hurt |
| move generic FrameNet-trained model from FrameNet test data to Communicator dialogs (ablation) | OnePar generic · lenient argument labeling | F-measure: 0.44 -> 0.25 | Hurt |
| compare OnePar generic model with SEMAFOR out of the box | OnePar vs SEMAFOR · Communicator lenient argument labeling | F-measure: 0.25 -> 0.24 | No clear effect |
Closest papers by meaning
- Fine-Grained Semantic Conceptualization of FrameNet (AAAI 2016 · similarity 0.86)
- Dependency Tree Representations of Predicate-Argument Structures (AAAI 2016 · similarity 0.85)
- Semantic Parsing in Task-Oriented Dialog with Recursive Insertion-Based Encoder (AAAI 2022 · similarity 0.85)
- Frame Semantic Role Labeling Using Arbitrary-Order Conditional Random Fields (AAAI 2024 · similarity 0.85)
- A Lens into Interpretable Transformer Mistakes via Semantic Dependency (ICML 2025 · similarity 0.84)
- Structure-Aware Multimodal Sequential Learning for Visual Dialog (AAAI 2024 · similarity 0.83)
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.