Research map / Vision-language and multimodal / Human and embodied modeling
Joint Semantic Analysis with Document-Level Cross-Task Coherence Rewards
Rahul Aralikatte, Mostafa Abdou, Heather C Lent, Daniel Hershcovich, Anders Søgaard
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| Fine-tune with document-level coherence rewards | Coreference resolver with LSTM + CNN · Six coreference datasets | mean F1: 49.01 -> 49.40 | Improved |
| Fine-tune with document-level coherence rewards | Coreference resolver with BERT-Tiny · Six coreference datasets | mean F1: 49.70 -> 50.95 | Improved |
| Fine-tune with document-level coherence rewards | Coreference resolver with BERT-Mini · Six coreference datasets | mean F1: 52.61 -> 52.88 | Improved |
| Fine-tune with document-level coherence rewards | Coreference resolver with BERT-Small · Six coreference datasets | mean F1: 52.76 -> 53.90 | Improved |
| Fine-tune with document-level coherence rewards | Coreference resolver with BERT-Medium · Six coreference datasets | mean F1: 55.67 -> 56.19 | Improved |
| Fine-tune with document-level coherence rewards | Coreference resolver with BERT-Base · Six coreference datasets | mean F1: 57.78 -> 58.18 | Improved |
| Fine-tune with document-level coherence rewards | SRL model with LSTM + CNN · Four SRL datasets | mean F1: 67.63 -> 67.74 | Improved |
| Fine-tune with document-level coherence rewards | SRL model with BERT-Tiny · Four SRL datasets | mean F1: 56.87 -> 57.08 | Improved |
| Fine-tune with document-level coherence rewards | SRL model with BERT-Mini · Four SRL datasets | mean F1: 70.51 -> 70.48 | Hurt |
| Fine-tune with document-level coherence rewards | SRL model with BERT-Small · Four SRL datasets | mean F1: 74.26 -> 74.48 | Improved |
| Fine-tune with document-level coherence rewards | SRL model with BERT-Medium · Four SRL datasets | mean F1: 75.62 -> 75.57 | Hurt |
| Fine-tune with document-level coherence rewards | SRL model with BERT-Base · Four SRL datasets | mean F1: 79.46 -> 79.52 | Improved |
Closest papers by meaning
- Have We Solved The Hard Problem? It’s Not Easy! Contextual Lexical Contrast as a Means to Probe Neural Coherence (AAAI 2021 · similarity 0.86)
- Deep Semantic Role Labeling With Self-Attention (AAAI 2018 · similarity 0.85)
- Coreference Graph Guidance for Mind-Map Generation (AAAI 2024 · similarity 0.85)
- Frame Semantic Role Labeling Using Arbitrary-Order Conditional Random Fields (AAAI 2024 · similarity 0.85)
- Syntax-Aware Neural Semantic Role Labeling (AAAI 2019 · similarity 0.85)
- BBScore: A Brownian Bridge Based Metric for Assessing Text Coherence (AAAI 2024 · similarity 0.84)
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.
