Research map / Label-efficient and robust learning / Recognition and long-tail
Knowing What, How and Why: A Near Complete Solution for Aspect-Based Sentiment Analysis
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 the two-stage aspect-sentiment-opinion extraction framework | ASTE · 14res, 14lap, 15res, 16res | triplet F1: 51.89 vs 51.68; 43.50 vs 42.47; 46.79 vs 46.69; 53.62 vs 44.51 | Helped |
| Use the two-stage framework for aspect-opinion pairing | ASTE · 14res, 14lap, 16res | pair F1: 56.10 vs 55.34; 53.85 vs 52.56; 60.04 vs 53.75 | Helped |
| Use the two-stage framework for aspect-opinion pairing | ASTE · 15res | pair F1: 56.23 vs 56.85 | Hurt |
| Add explicit opinion extraction to Li-unified | Li-unified aspect-sentiment model · 14res, 14lap, 15res, 16res | aspect-sentiment F1: 71.75 -> 73.79; 61.86 -> 63.38; 64.52 -> 64.95; 69.06 -> 70.20 | Helped |
| Add the opinion BLSTM layer (ablation) | Our · 14res, 14lap, 15res, 16res | aspect-sentiment F1: 72.04 -> 71.95; 59.77 -> 62.34; 64.29 -> 65.79; 70.67 -> 71.73 | Mixed |
| Add target guidance to exchange target information with opinion extraction (ablation) | Our · 14res, 14lap, 15res, 16res | aspect-sentiment F1: 74.19 -> 71.95; 62.26 -> 62.34; 60.57 -> 65.79; 65.71 -> 71.73 | Helped |
| Add target guidance (ablation) | Our · 14res, 14lap | opinion extraction F1: 83.21 -> 82.45; 76.03 -> 74.84 | Hurt |
| Add target guidance (ablation) | Our · 15res, 16res | opinion extraction F1: 76.10 -> 78.02; 83.67 -> 83.73 | Helped |
| Add target-boundary auxiliary loss (ablation) | Our · 14res, 14lap, 15res, 16res | aspect-sentiment F1: 70.79 -> 71.95; 62.34 -> 62.34; 64.25 -> 65.79; 66.74 -> 71.73 | Mixed |
| Add target-boundary auxiliary loss (ablation) | Our · 14res, 14lap, 15res, 16res | opinion extraction F1: 82.88 -> 82.45; 75.21 -> 74.84; 76.60 -> 78.02; 81.77 -> 83.73 | Mixed |
| Constrain each aspect to a fixed number of predicted triplets | Stage-two pairing · ASTE datasets | triplet extraction | No clear effect |
Closest papers by meaning
- Aspect-Based Sentiment Analysis with Explicit Sentiment Augmentations (AAAI 2024 · similarity 0.92)
- Aspect Enhancement and Text Simplification in Multimodal Aspect-Based Sentiment Analysis for Multi-Aspect and Multi-Sentiment Scenarios (AAAI 2025 · similarity 0.88)
- DanceHA: A Multi-Agent Framework for Document-Level Aspect-Based Sentiment Analysis (AAAI 2026 · similarity 0.87)
- TextGT: A Double-View Graph Transformer on Text for Aspect-Based Sentiment Analysis (AAAI 2024 · similarity 0.86)
- A Novel Energy Based Model Mechanism for Multi-Modal Aspect-Based Sentiment Analysis (AAAI 2024 · similarity 0.86)
- You Only Read Once: Constituency-Oriented Relational Graph Convolutional Network for Multi-Aspect Multi-Sentiment Classification (AAAI 2024 · similarity 0.86)
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.