Papers by Sohyun Park
Label-aware Hard Negative Sampling Strategies with Momentum Contrastive Learning for Implicit Hate Speech Detection (2024.findings-acl)
Copied to clipboard
| Challenge: | Existing models for implicit hate speech detection do not have significant advantage over cross-entropy loss-based learning. |
| Approach: | They propose a label-aware hard negative sampling strategy that encourages the model to learn detailed features from hard negative samples instead of random batch. |
| Outcome: | The proposed models outperform existing models for implicit hate speech detection both in- and cross-datasets. |
PREDICT: Multi-Agent-based Debate Simulation for Generalized Hate Speech Detection (2024.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods to generalize hate speech detection models have been limited by the labeling criteria between datasets. |
| Approach: | They propose a framework that uses the concept of multi-agent for hate speech detection that uses a set of labeling criteria to create multiple agents based on the induced labeling of given datasets. |
| Outcome: | The proposed framework achieves superior cross-evaluation performance compared to methods that focus on specific labeling criteria or majority voting methods. |