Papers by Lukas Christ
Training-Free Text Emotion Tagging via LLM-Based Best-Worst Scaling (2026.findings-eacl)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) are often used as automatic annotators for tasks such as Text Emotion Recognition (TER). |
| Approach: | They propose a novel algorithm that leverages Best-Worst Scaling to prompt the LLM to choose the least and most suitable emotions for a given text from several label subsets. |
| Outcome: | The proposed method compares favorably to existing methods and naive prompting approaches in terms of accuracy and calibration. |
Modeling Emotional Trajectories in Written Stories Utilizing Transformers and Weakly-Supervised Learning (2024.findings-acl)
Copied to clipboard
| Challenge: | Existing work on how to model emotionality in stories has been limited to dictionary-based methods . |
| Approach: | They propose to introduce continuous valence and arousal labels for an existing dataset of children’s stories originally annotated with discrete emotion categories. |
| Outcome: | The proposed model achieves a Concordance Correlation Coefficient (CCC) of .8221 for valence and .7125 for arousal on the test set, demonstrating the efficacy of the proposed model. |