Research map / Graph learning / Graph self-supervised learning
Neural Sentence Ordering Based on Constraint Graphs
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 |
|---|---|---|---|
| Use constraint graphs at multiple distances rather than a single graph (ablation) | Sentence ordering model · AAN abstracts and SIND captions | Kendall's tau: AAN: 0.8078 -> 0.8236; SIND: 0.5708 -> 0.5856 | Helped |
| Fine-tune the second-phase BERT encoder (ablation) | Sentence ordering model · AAN abstracts and SIND captions | Kendall's tau: AAN: 0.8236 -> 0.8245; SIND: 0.5856 -> 0.5867 | Helped |
| Replace BERT with RoBERTa (ablation) | Sentence ordering model · AAN abstracts and SIND captions | Kendall's tau: AAN: 0.8236 -> 0.8240; SIND: 0.5856 -> 0.5761 | Mixed |
| Replace BERT with ALBERT (ablation) | Sentence ordering model · AAN abstracts and SIND captions | Kendall's tau: AAN: 0.8236 -> 0.8228; SIND: 0.5856 -> 0.5820 | Mixed |
| Replace GIN with GCN (ablation) | Sentence ordering model · AAN abstracts and SIND captions | Kendall's tau: AAN: 0.8236 -> 0.8206; SIND: 0.5856 -> 0.5818 | Hurt |
| Use only g2 instead of all three graphs (ablation) | Sentence ordering model · AAN abstracts and SIND captions | Kendall's tau: AAN: 0.8170 vs 0.8236; SIND: 0.5377 vs 0.5856 | Hurt |
| Use only g3 instead of all three graphs (ablation) | Sentence ordering model · AAN abstracts and SIND captions | Kendall's tau: AAN: 0.8116 vs 0.8236; SIND: 0.5276 vs 0.5856 | Hurt |
| Use g1+g2 instead of all three graphs (ablation) | Sentence ordering model · AAN abstracts and SIND captions | Kendall's tau: AAN: 0.8217 vs 0.8236; SIND: 0.5744 vs 0.5856 | Hurt |
| Use g1+g3 instead of all three graphs (ablation) | Sentence ordering model · AAN abstracts and SIND captions | Kendall's tau: AAN: 0.8196 vs 0.8236; SIND: 0.5837 vs 0.5856 | Hurt |
| Use g2+g3 instead of all three graphs (ablation) | Sentence ordering model · AAN abstracts and SIND captions | Kendall's tau: AAN: 0.8208 vs 0.8236; SIND: 0.5539 vs 0.5856 | Hurt |
Closest papers by meaning
- Enhancing Pointer Network for Sentence Ordering with Pairwise Ordering Predictions (AAAI 2020 · similarity 0.88)
- Sentence Ordering and Coherence Modeling using Recurrent Neural Networks (AAAI 2018 · similarity 0.87)
- Deep Attentive Ranking Networks for Learning to Order Sentences (AAAI 2020 · similarity 0.86)
- Implicit Word Reordering with Knowledge Distillation for Cross-Lingual Dependency Parsing (AAAI 2025 · similarity 0.83)
- Order-Independence Without Fine Tuning (NeurIPS 2024 · similarity 0.82)
- Narrowing the Gap between Supervised and Unsupervised Sentence Representation Learning with Large Language Model (AAAI 2024 · similarity 0.82)
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.