Research map / Label-efficient and robust learning / Recognition and long-tail
Unity in Diversity: Learning Distributed Heterogeneous Sentence Representation for Extractive Summarization
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 |
|---|---|---|---|
| pretrain HNet on Daily Mail, then fine-tune on DUC (ablation) | HNet · DUC-2001, DUC-2002, DUC-2004 | ROUGE-1 and ROUGE-2: HNet: 37.21/8.96, 39.17/9.61, 39.54/10.94 -> HNet(T): 38.18/9.43, 39.94/9.92, 40.34/11.29 | Helped |
| use CSTI with BTI instead of the Convolutional BLSTM encoder (ablation) | HNet · DUC-2001, DUC-2002, DUC-2004 | ROUGE-1 and ROUGE-2: HNet vs HNet-B: 37.21/8.96 vs 36.82/8.64; 39.17/9.61 vs 38.79/9.43; 39.54/10.94 vs 39.27/10.85 | Helped |
| add document-dependent features to the sentence representation (ablation) | HNet⁻ · DUC-2001, DUC-2002, DUC-2004 | ROUGE-1 and ROUGE-2: HNet⁻ -> HNet: 34.51/7.88 -> 37.21/8.96; 35.86/8.24 -> 39.17/9.61; 35.66/9.37 -> 39.54/10.94 | Helped |
| use HNet with CSTI rather than HNet-B with Convolutional BLSTM (ablation) | HNet · DUC-2001, DUC-2002, DUC-2004 | ROUGE-1 and ROUGE-2: HNet-B(T) -> HNet(T): 37.69/9.12 -> 38.18/9.43; 39.52/9.69 -> 39.94/9.92; 39.9/11.08 -> 40.34/11.29 | Helped |
Closest papers by meaning
- SummaRuNNer: A Recurrent Neural Network Based Sequence Model for Extractive Summarization of Documents (AAAI 2017 · similarity 0.88)
- Keywords-Guided Abstractive Sentence Summarization (AAAI 2020 · similarity 0.88)
- Hierarchical Heterogeneous Graph Attention Network for Syntax-Aware Summarization (AAAI 2022 · similarity 0.88)
- Summarization Programs: Interpretable Abstractive Summarization with Neural Modular Trees (ICLR 2023 · similarity 0.88)
- Unsupervised Extractive Summarization with Learnable Length Control Strategies (AAAI 2024 · similarity 0.87)
- Language Models of Code Are Few-Shot Planners and Reasoners for Multi-Document Summarization with Attribution (AAAI 2025 · similarity 0.85)
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.