TwiUSD: A Benchmark Dataset and Structure-Aware LLM Framework for User Stance Detection

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Niu, Fuqiang, Chen, Zini, Xie, Zhiyu, Huang, Hu, Dai, Genan, Zhang, Bowen
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909700949278720
author Niu, Fuqiang
Chen, Zini
Xie, Zhiyu
Huang, Hu
Dai, Genan
Zhang, Bowen
author_facet Niu, Fuqiang
Chen, Zini
Xie, Zhiyu
Huang, Hu
Dai, Genan
Zhang, Bowen
contents User-level stance detection (UserSD) remains challenging due to the lack of high-quality benchmarks that jointly capture linguistic and social structure. In this paper, we introduce TwiUSD, the first large-scale, manually annotated UserSD benchmark with explicit followee relationships, containing 16,211 users and 47,757 tweets. TwiUSD enables rigorous evaluation of stance models by integrating tweet content and social links, with superior scale and annotation quality. Building on this resource, we propose MRFG: a structure-aware framework that uses LLM-based relevance filtering and feature routing to address noise and context heterogeneity. MRFG employs multi-scale filtering and adaptively routes features through graph neural networks or multi-layer perceptrons based on topological informativeness. Experiments show MRFG consistently outperforms strong baselines (including PLMs, graph-based models, and LLM prompting) in both in-target and cross-target evaluation.
format Preprint
id arxiv_https___arxiv_org_abs_2506_13343
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle TwiUSD: A Benchmark Dataset and Structure-Aware LLM Framework for User Stance Detection
Niu, Fuqiang
Chen, Zini
Xie, Zhiyu
Huang, Hu
Dai, Genan
Zhang, Bowen
Social and Information Networks
User-level stance detection (UserSD) remains challenging due to the lack of high-quality benchmarks that jointly capture linguistic and social structure. In this paper, we introduce TwiUSD, the first large-scale, manually annotated UserSD benchmark with explicit followee relationships, containing 16,211 users and 47,757 tweets. TwiUSD enables rigorous evaluation of stance models by integrating tweet content and social links, with superior scale and annotation quality. Building on this resource, we propose MRFG: a structure-aware framework that uses LLM-based relevance filtering and feature routing to address noise and context heterogeneity. MRFG employs multi-scale filtering and adaptively routes features through graph neural networks or multi-layer perceptrons based on topological informativeness. Experiments show MRFG consistently outperforms strong baselines (including PLMs, graph-based models, and LLM prompting) in both in-target and cross-target evaluation.
title TwiUSD: A Benchmark Dataset and Structure-Aware LLM Framework for User Stance Detection
topic Social and Information Networks
url https://arxiv.org/abs/2506.13343