Abstract
We present our system for SemEval-2026 Task 3 on dimensional aspect-based sentiment regression. Our approach combines a hybrid RoBERTa encoder, which jointly predicts sentiment using regression and discretized classification heads, with large language models (LLMs) via prediction-level ensemble learning. The hybrid encoder improves prediction stability by combining continuous and discretized sentiment representations. We further explore in-context learning with LLMs and ridge-regression stacking to combine encoder and LLM predictions. Experimental results on the development set show that ensemble learning significantly improves performance over individual models, achieving substantial reductions in RMSE and improvements in correlation scores. Our findings demonstrate the complementary strengths of encoder-based and LLM-based approaches for dimensional sentiment analysis. Our development code and resources will be shared at this https URL
Abstract (translated)
我们介绍了我们的系统,用于参加2026年SemEval竞赛的第三项任务:基于维度的方面情感回归。我们的方法结合了一种混合RoBERTa编码器与大型语言模型(LLMs),通过预测级集成学习将两者结合起来。该混合编码器通过整合连续和离散的情感表示来提高预测稳定性,从而联合进行回归和离散分类以预测情感。我们进一步探索了在情境中使用LLM的在上下文学习以及岭回归堆叠技术,以便结合编码器和LLM的预测结果。 实验结果显示,在开发集上,集成学习显著提高了单一模型的表现力,实现了RMSE(均方根误差)的大幅度降低,并且相关得分有所提升。我们的研究发现表明了基于编码器的方法与基于大型语言模型的方法在维度情感分析中的互补优势。我们将通过以下链接分享我们的开发代码和资源:[此URL]
URL
https://arxiv.org/abs/2603.07766