Improving Neural Argumentative Stance Classification in Controversial Topics with Emotion-Lexicon Features
arXiv:2602.22846v1 Announce Type: new Abstract: Argumentation mining comprises several subtasks, among which stance classification focuses on identifying the standpoint expressed in an argumentative text toward a specific target topic. While arguments-especially about controversial topics-often appeal to emotions, most prior work has not systematically incorporated explicit, fine-grained emotion analysis to improve performance on this task. In particular, prior research on stance classification has predominantly utilized non-argumentative texts and has been restricted to specific domains or topics, limiting generalizability. We work on five datasets from diverse domains encompassing a range of controversial topics and present an approach for expanding the Bias-Corrected NRC Emotion Lexicon using DistilBERT embeddings, which we feed into a Neural Argumentative Stance Classification model. Our method systematically expands the emotion lexicon through contextualized embeddings to identif
arXiv:2602.22846v1 Announce Type: new Abstract: Argumentation mining comprises several subtasks, among which stance classification focuses on identifying the standpoint expressed in an argumentative text toward a specific target topic. While arguments-especially about controversial topics-often appeal to emotions, most prior work has not systematically incorporated explicit, fine-grained emotion analysis to improve performance on this task. In particular, prior research on stance classification has predominantly utilized non-argumentative texts and has been restricted to specific domains or topics, limiting generalizability. We work on five datasets from diverse domains encompassing a range of controversial topics and present an approach for expanding the Bias-Corrected NRC Emotion Lexicon using DistilBERT embeddings, which we feed into a Neural Argumentative Stance Classification model. Our method systematically expands the emotion lexicon through contextualized embeddings to identify emotionally charged terms not previously captured in the lexicon. Our expanded NRC lexicon (eNRC) improves over the baseline across all five datasets (up to +6.2 percentage points in F1 score), outperforms the original NRC on four datasets (up to +3.0), and surpasses the LLM-based approach on nearly all corpora. We provide all resources-including eNRC, the adapted corpora, and model architecture-to enable other researchers to build upon our work.
Executive Summary
This study addresses the limitation of prior work in argumentative stance classification by incorporating explicit, fine-grained emotion analysis using a Bias-Corrected NRC Emotion Lexicon. The authors expand the lexicon using DistilBERT embeddings and feed it into a Neural Argumentative Stance Classification model. Their approach improves performance on five diverse datasets from controversial topics, outperforming baseline and LLM-based models. The study provides a valuable resource for researchers, expanding the emotion lexicon and improving stance classification. The findings have significant implications for natural language processing and argumentation mining applications, particularly in areas such as opinion mining, sentiment analysis, and debate analysis.
Key Points
- ▸ The study addresses a significant limitation in prior work on argumentative stance classification by incorporating emotion analysis.
- ▸ The authors expand the Bias-Corrected NRC Emotion Lexicon using DistilBERT embeddings and feed it into a Neural Argumentative Stance Classification model.
- ▸ The approach improves performance on five diverse datasets from controversial topics, outperforming baseline and LLM-based models.
Merits
Improved Performance
The study demonstrates significant improvements in stance classification performance using the expanded emotion lexicon, particularly on datasets with diverse and controversial topics.
Generalizability
The approach is generalizable across different domains and topics, increasing its applicability in real-world applications.
Demerits
Limited Evaluation Datasets
The study is limited to five datasets, and it is unclear whether the approach will generalize to other, potentially more challenging, evaluation datasets.
Dependence on Pre-trained Models
The approach relies heavily on pre-trained models (e.g., DistilBERT), which may limit its portability to other datasets or domains.
Expert Commentary
The study is a significant contribution to the field of argumentation mining, as it addresses a long-standing limitation in prior work by incorporating explicit, fine-grained emotion analysis. The use of DistilBERT embeddings to expand the emotion lexicon is a valuable innovation, and the improved performance on diverse datasets is impressive. However, the study is limited by its reliance on pre-trained models and the lack of evaluation on more challenging datasets. Nevertheless, the study provides a valuable resource for researchers and practitioners working on argumentation mining and natural language processing applications.
Recommendations
- ✓ Future studies should investigate the generalizability of the approach to other, potentially more challenging, evaluation datasets.
- ✓ Researchers should explore the use of other pre-trained models or techniques to improve the portability of the approach to different domains and topics.
