Research map / Label-efficient and robust learning / Recognition and long-tail
Unsupervised Learning of Discourse Structures using a Tree Autoencoder
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 |
|---|---|---|---|
| Train T-AE on the larger Yelp'13 corpus instead of RST-DT | T-AE · RST-DT | average micro-precision: 69.68 -> 71.32 | Helped |
| Train T-AE on Yelp'13 instead of RST-DT | T-AE document-level encodings · Yelp'13 | accuracy: 40.41 -> 42.69 | Helped |
| Use MSE rather than KLD as the reconstruction loss (ablation) | T-AE · development set | performance | Helped |
Closest papers by meaning
- Predicting Above-Sentence Discourse Structure Using Distant Supervision from Topic Segmentation (AAAI 2022 · similarity 0.90)
- A Deep Sequential Model for Discourse Parsing on Multi-Party Dialogues (AAAI 2019 · similarity 0.89)
- Tree-Structured Neural Machine for Linguistics-Aware Sentence Generation (AAAI 2018 · similarity 0.88)
- Tree-Structured Attention with Hierarchical Accumulation (ICLR 2020 · similarity 0.87)
- Improving Unsupervised Constituency Parsing via Maximizing Semantic Information (ICLR 2025 · similarity 0.84)
- Ensemble Distillation for Unsupervised Constituency Parsing (ICLR 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.