Research map / Label-efficient and robust learning / Recognition and long-tail
Weakly-Supervised Fine-Grained Event Recognition on Social Media Texts for Disaster Management
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 |
|---|---|---|---|
| Clean keyword-identified training tweets with clustering-assisted word-sense disambiguation (ablation) | Basic event classifier · Hurricane Harvey | Macro-average F1: 54.9 -> 62.3 | Helped |
| Clean keyword-identified training tweets with clustering-assisted word-sense disambiguation (ablation) | Basic event classifier · Hurricane Florence | Macro-average F1: 44.1 -> 57.3 | Helped |
| Add context and reply tweet channels (ablation) | Basic classifier trained on cleaned seed tweets · Hurricane Harvey | Macro-average F1: 62.3 -> 63.8 | Helped |
| Add context and reply tweet channels (ablation) | Basic classifier trained on cleaned seed tweets · Hurricane Florence | Macro-average F1: 57.3 -> 58.5 | Helped |
| Add bootstrapping (ablation) | Basic classifier trained on cleaned seed tweets · Hurricane Harvey | Macro-average F1: 62.3 -> 63.3 | Helped |
| Add bootstrapping (ablation) | Multi-channel classifier trained on cleaned seed tweets · Hurricane Harvey | Macro-average F1: 63.8 -> 65.5 | Helped |
| Add bootstrapping (ablation) | Basic classifier trained on cleaned seed tweets · Hurricane Florence | Macro-average F1: 57.3 -> 58.8 | Helped |
| Add bootstrapping (ablation) | Multi-channel classifier trained on cleaned seed tweets · Hurricane Florence | Macro-average F1: 58.5 -> 59.7 | Helped |
| Add bootstrapping (ablation) | Multi-channel classifier trained on uncleaned seed tweets · Hurricane Harvey | Macro-average F1: 56.5 -> 56.4 | No clear effect |
| Use clustering-assisted WSD-cleaned seed tweets rather than uncleaned keyword tweets (ablation) | Multi-channel classifier + bootstrapping · Hurricane Harvey | Macro-average F1: 56.4 -> 65.5 | Helped |
| Use the weakly supervised full system rather than the supervised classifier | Hurricane Harvey event recognition · Hurricane Harvey | Macro-average F1: 61.9 -> 65.5 | Helped |
| Use the weakly supervised full system rather than the supervised classifier | Hurricane Florence event recognition · Hurricane Florence | Macro-average F1: 47.9 -> 59.7 | Helped |
Closest papers by meaning
- Unsupervised Detection of Sub-Events in Large Scale Disasters (AAAI 2020 · similarity 0.91)
- Towards Effective, Efficient and Unsupervised Social Event Detection in the Hyperbolic Space (AAAI 2025 · similarity 0.84)
- Hierarchical and Incremental Structural Entropy Minimization for Unsupervised Social Event Detection (AAAI 2024 · similarity 0.82)
- MONITRS: Multimodal Observations of Natural Incidents Through Remote Sensing (NeurIPS 2025 · similarity 0.81)
- EventSum: A Large-Scale Event-Centric Summarization Dataset for Chinese Multi-News Documents (AAAI 2025 · similarity 0.81)
- Beyond Grounding: Extracting Fine-Grained Event Hierarchies across Modalities (AAAI 2024 · similarity 0.81)
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.