Papers by Zhechao Zhu
Pinpointing Diffusion Grid Noise to Enhance Aspect Sentiment Quad Prediction (2024.findings-acl)
Copied to clipboard
| Challenge: | Current studies on aspect-based sentiment analysis focus on essential content for model generation, ignoring the incorporation of various noise during training. |
| Approach: | They propose a grid noise-diffusion pinpoint network (GDP) model that incorporates three new modules to tackle generation instability. |
| Outcome: | The proposed model reduces the generation instability of model learning and outputs by incorporating Consistency Likelihood Learning and GDP-FOR. |