# Incorporating Count-Based Features into Pre-Trained Models for Improved Stance Detection

arXiv: https://arxiv.org/abs/2010.09078 · HTML: https://mlsanity.com/paper/2010.09078/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 18 Oct 2020.

## Built on by (most cited first, 6 of 6)

- [POLITICS: Pretraining with Same-story Article Comparison for Ideology Prediction and Stance Detection](https://mlsanity.com/paper/2205.00619/) (2022)
- [PAR: Political Actor Representation Learning with Social Context and Expert Knowledge](https://mlsanity.com/paper/2210.08362/) (2022)
- [Explaining Translationese: why are Neural Classifiers Better and what do they Learn?](https://mlsanity.com/paper/2210.13391/) (2022)
- [UoB at SemEval-2021 Task 5: Extending Pre-Trained Language Models to Include Task and Domain-Specific Information for Toxic Span Prediction](https://mlsanity.com/paper/2110.03730/) (2021)
- [Emotion-aware Dual Cross-Attentive Neural Network with Label Fusion for Stance Detection in Misinformative Social Media Content](https://mlsanity.com/paper/2505.23812/) (2025)
- [Toward Effective Multi-Domain Rumor Detection in Social Networks Using Domain-Gated Mixture-of-Experts](https://mlsanity.com/paper/2602.21214/) (2026)

## Abstract (arXiv)

The explosive growth and popularity of Social Media has revolutionised the way we communicate and collaborate. Unfortunately, this same ease of accessing and sharing information has led to an explosion of misinformation and propaganda. Given that stance detection can significantly aid in veracity prediction, this work focuses on boosting automated stance detection, a task on which pre-trained models have been extremely successful on, as on several other tasks. This work shows that the task of stance detection can benefit from feature based information, especially on certain under performing classes, however, integrating such features into pre-trained models using ensembling is challenging. We propose a novel architecture for integrating features with pre-trained models that address these challenges and test our method on the RumourEval 2019 dataset. This method achieves state-of-the-art results with an F1-score of 63.94 on the test set.