Saved in:
Bibliographic Details
Main Authors: Huang, Liyuan, He, Jiawei, Shen, Wutao, Li, Lin, Zhang, Jin
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.10560
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917480642904064
author Huang, Liyuan
He, Jiawei
Shen, Wutao
Li, Lin
Zhang, Jin
author_facet Huang, Liyuan
He, Jiawei
Shen, Wutao
Li, Lin
Zhang, Jin
contents This paper describes our system to SemEval-2026 Task 3 Track A Subtask 1 on Dimensional Aspect Sentiment Regression (DimASR). We propose a lightweight and resource-efficient system built entirely on multilingual pre-trained encoders, without relying on LLMs or external corpora. We adopt joint multilingual and multi-domain training to facilitate cross-lingual transfer and alleviate data sparsity, introduce a bounded regression transformation that improves training stability while constraining predictions within the valid range, and employ an adaptive ensemble strategy via subset search to reduce prediction variance. Experimental results demonstrate that our system achieves strong and consistent performance, ranking 1st on zho-res, 2nd on zho-lap, and 3rd on jpn-hot, with all remaining datasets placed within the top half of participating teams.
format Preprint
id arxiv_https___arxiv_org_abs_2605_10560
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle ICT-NLP at SemEval-2026 Task 3: Less Is More -- Multilingual Encoder with Joint Training and Adaptive Ensemble for Dimensional Aspect Sentiment Regression
Huang, Liyuan
He, Jiawei
Shen, Wutao
Li, Lin
Zhang, Jin
Computation and Language
This paper describes our system to SemEval-2026 Task 3 Track A Subtask 1 on Dimensional Aspect Sentiment Regression (DimASR). We propose a lightweight and resource-efficient system built entirely on multilingual pre-trained encoders, without relying on LLMs or external corpora. We adopt joint multilingual and multi-domain training to facilitate cross-lingual transfer and alleviate data sparsity, introduce a bounded regression transformation that improves training stability while constraining predictions within the valid range, and employ an adaptive ensemble strategy via subset search to reduce prediction variance. Experimental results demonstrate that our system achieves strong and consistent performance, ranking 1st on zho-res, 2nd on zho-lap, and 3rd on jpn-hot, with all remaining datasets placed within the top half of participating teams.
title ICT-NLP at SemEval-2026 Task 3: Less Is More -- Multilingual Encoder with Joint Training and Adaptive Ensemble for Dimensional Aspect Sentiment Regression
topic Computation and Language
url https://arxiv.org/abs/2605.10560