Research map / Architectures and efficiency / Attention and transformers
Be Relevant, Non-Redundant, and Timely: Deep Reinforcement Learning for Real-Time Event Summarization
Min Yang, Chengming Li, Fei Sun, Zhou Zhao, Ying Shen, Chenglin Wu
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 |
|---|---|---|---|
| add reinforcement learning (ablation) | DRES · TREC-RTS-16 | GMP: -0.102 -> -0.080 | Improved |
| add reinforcement learning (ablation) | DRES · TREC-RTS-17 | GMP: -0.094 -> -0.056 | Improved |
| add relevance prediction auxiliary task (ablation) | DRES · TREC-RTS-16 | GMP: -0.097 -> -0.080 | Improved |
| add relevance prediction auxiliary task (ablation) | DRES · TREC-RTS-17 | GMP: -0.097 -> -0.056 | Improved |
| add multi-topic dynamic memory network (ablation) | DRES · TREC-RTS-16 | GMP: -0.091 -> -0.080 | Improved |
| add multi-topic dynamic memory network (ablation) | DRES · TREC-RTS-17 | GMP: -0.067 -> -0.056 | Improved |
| add context-aware Transformer (ablation) | DRES · TREC-RTS-16 | EG-1: 0.299 -> 0.308 | Improved |
| add multi-perspective cross-attention (ablation) | DRES · TREC-RTS-17 | EG-1: 0.295 -> 0.302 | Improved |
| add reinforcement learning (ablation) | DRES · TREC-RTS-16 | EG-0: 0.071 -> 0.087 | Improved |
| add reinforcement learning (ablation) | DRES · TREC-RTS-17 | Latency: 40182 -> 35319 | Improved |
| replace representation and Transformer layers with pre-trained BERT (DRES+Bert) | DRES · TREC-RTS-16 | EG-1: 0.308 -> 0.306 | No clear effect |
| replace representation and Transformer layers with pre-trained BERT (DRES+Bert) | DRES · TREC-RTS-17 | EG-1: 0.302 -> 0.298 | No clear effect |
Closest papers by meaning
- A Deep Reinforced Model for Abstractive Summarization (ICLR 2018 · similarity 0.88)
- EventSum: A Large-Scale Event-Centric Summarization Dataset for Chinese Multi-News Documents (AAAI 2025 · similarity 0.84)
- CAR-Transformer: Cross-Attention Reinforcement Transformer for Cross-Lingual Summarization (AAAI 2024 · similarity 0.83)
- Learning to Summarize by Learning to Quiz: Adversarial Agentic Collaboration for Long Document Summarization (ICLR 2026 · similarity 0.83)
- Document Summarization with Conformal Importance Guarantees (NeurIPS 2025 · similarity 0.83)
- Unsupervised Extractive Summarization with Learnable Length Control Strategies (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.
