Papers by Yi-Shin Chen
ConTextING: Granting Document-Wise Contextual Embeddings to Graph Neural Networks for Inductive Text Classification (2022.coling-1)
Copied to clipboard
| Challenge: | Graph neural networks (GNNs) are used to learn document representation from graph structures. |
| Approach: | They propose a unified model with a joint training mechanism to learn from document embeddings and contextual word interactions simultaneously. |
| Outcome: | The proposed model outperforms pure inductive GNNs and BERT-style models . the proposed model also has a joint training mechanism to learn from document embeddings and contextual word interactions simultaneously. |
Leveraging Conflicts in Social Media Posts: Unintended Offense Dataset (2024.emnlp-main)
Copied to clipboard
| Challenge: | a new study examines the impact of conflict on multi-person communication datasets on offensive language . conflict datasets often neglect contextual information and focus on intended offenses . authors propose a conflict-based data collection method to analyze inter-conflict cues in multi-user communications . |
| Approach: | They propose a conflict-based data collection method to utilize inter-conflict cues in multi-person communications. |
| Outcome: | The proposed method improves the accuracy of detecting offensive language and enriches our understanding of conflict dynamics in digital communication. |
CARER: Contextualized Affect Representations for Emotion Recognition (D18-1)
Copied to clipboard
| Challenge: | Existing methods to model emotion-relevant content are based on rule-based and statistics-based approaches. |
| Approach: | They propose a semi-supervised graph-based algorithm to produce rich structural descriptors . they use word embeddings to evaluate the algorithm on emotion recognition tasks . |
| Outcome: | The proposed method outperforms state-of-the-art methods on emotion recognition tasks. |