What Matters in Irony Detection: An Extended Feature Engineering for Irony Detection in English Tweets.
In recent years, large-scale language models (LLMs) have nearly become the dominant force in almost every natural language processing (NLP) task. The primary research approach has focused on selecting the most appropriate language model for specific NLP tasks and then incorporating linguistic featur...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
LibraryPress@UF
2024-05-01
|
| Series: | Proceedings of the International Florida Artificial Intelligence Research Society Conference |
| Online Access: | https://journals.flvc.org/FLAIRS/article/view/135276 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | In recent years, large-scale language models (LLMs) have nearly become the dominant force in almost every natural language processing (NLP) task. The primary research approach has focused on selecting the most appropriate language model for specific NLP tasks and then incorporating linguistic features to enhance the model’s performance. With swift progress in this field, new features and models are evolving rapidly, and outdated systems require timely updates. In this paper, we extended the accomplishments of SemEval-2018 Task 3, enhancing its irony detection systems with novel features and more sophisticated language models. Subsequently, we conducted an ablation study to showcase the contributions of these enhancements to the LLM-based system. Furthermore, we compared our leading system with the top performers in the SemEval-2018 competition, and our best model exhibited superior performance when compared to the leading performers applied to the same corpus. |
|---|---|
| ISSN: | 2334-0754 2334-0762 |