Research map / Architectures and efficiency / Attention and transformers
SummaRuNNer: A Recurrent Neural Network Based Sequence Model for Extractive Summarization of Documents
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 extractive rather than abstractive training (ablation) | SummaRuNNer · Daily Mail test, 75 bytes | Rouge-1 recall: 23.8 -> 26.2$\pm 0.4$ | Helped |
| Use extractive rather than abstractive training (ablation) | SummaRuNNer · Daily Mail test, 75 bytes | Rouge-2 recall: 9.6 -> 10.8$\pm 0.3$ | Helped |
| Use extractive rather than abstractive training (ablation) | SummaRuNNer · Daily Mail test, 75 bytes | Rouge-L recall: 13.3 -> 14.4$\pm 0.3$ | Helped |
| Use extractive rather than abstractive training (ablation) | SummaRuNNer · Daily Mail test, 275 bytes | Rouge-1 recall: 40.4 -> 42.0 $\pm 0.2$ | Helped |
| Use extractive rather than abstractive training (ablation) | SummaRuNNer · CNN/Daily Mail test | Rouge-1 F1: 37.5 -> 39.6$\pm0.2$ | Helped |
| Use extractive rather than abstractive training (ablation) | SummaRuNNer · DUC 2002, 75 words | Rouge-1 recall: 44.8 -> 46.6 $\pm 0.8$ | Helped |
| Use SummaRuNNer instead of Cheng et al. '16 | Daily Mail test, 275 bytes · Daily Mail test, 275 bytes | Rouge-1 recall: 42.2 vs 42.0 $\pm 0.2$ | No clear effect |
| Use SummaRuNNer instead of Lead-3 | CNN/Daily Mail test · CNN/Daily Mail test | Rouge-L F1: 35.5 vs 35.3$\pm0.2$ | No clear effect |
| Use SummaRuNNer instead of Cheng et al. '16 | DUC 2002, 75 words · DUC 2002, 75 words | Rouge-L recall: 43.5 vs 43.03 $\pm 0.8$ | No clear effect |
Closest papers by meaning
- A Deep Reinforced Model for Abstractive Summarization (ICLR 2018 · similarity 0.89)
- Structured Neural Summarization (ICLR 2019 · similarity 0.89)
- Unity in Diversity: Learning Distributed Heterogeneous Sentence Representation for Extractive Summarization (AAAI 2018 · similarity 0.88)
- Text Summarization with Oracle Expectation (ICLR 2023 · similarity 0.88)
- MeanSum: A Neural Model for Unsupervised Multi-Document Abstractive Summarization (ICML 2019 · similarity 0.88)
- Unsupervised Extractive Summarization with Learnable Length Control Strategies (AAAI 2024 · similarity 0.88)
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.