Papers with data fitting
SenticVec: Toward Robust and Human-Centric Neurosymbolic Sentiment Analysis (2024.findings-acl)
Copied to clipboard
| Challenge: | Existing approaches to tackle learning challenges such as knowledge forgetting and extensive computing resources are not effective. |
| Approach: | They propose a novel neurosymbolic method for sentiment analysis that places emphasis on human subjectivity within varying domain annotations. |
| Outcome: | The proposed method is lightweight, robust across domains and languages, efficient few-shot training, and rapid convergence. |