arXiv cs.CL· Nishan Chatterjee, Matthew Purver, Antoine Doucet, Senja Pollak·· 11 小时前AI 评分20
仅找出错误还不够：面向实体情感分析的选择性模型升级策略
Finding Errors Is Not Enough: Selective Model Escalation for Entity-Targeted Sentiment Analysis
AI 导读
研究在文档级实体级情感分析（ELSA）中，用不确定性决定哪些预测需要复核、处理多少条，以及能否由替代编码器或本地 LLM 纠正。在 SentiCoref 上，若完美纠正最不确定 10% 的全部错误可提升 9.14 Macro-F1，但最佳实际动作仅增益 0.90。
正文 · 原文
Abstract:Improving a sentiment classifier with limited resources involves more than simply deciding how much data to collect or which model to train. A labelled collection may lack examples of an important class. Once trained, a classifier may benefit from another model even when that model is weaker on its own. We study these decisions in document-level Entity-Level Sentiment Analysis (ELSA), primarily using South Slavic news. We first compare acquisition orders as collections grow. We then use uncertainty to choose which predictions to revisit, how many to process, and whether an alternative encoder or local LLM can correct them. Finally, we select complete routing policies on development data for held-out evaluation. Minority-prioritised acquisition helps at some very small training sizes, but not consistently thereafter. On SentiCoref, an oracle correcting all errors in the least-confident 10% without harming correct predictions would add 9.14 Macro-F1 points; however the best evaluated action gains only 0.90. Encoder averaging helps on AspectBench and NewsMTSC but harms SentiCoref, where selective LLM assistance is more promising. Both development-selected policies route 30% of records to the encoder-assisted LLM, yielding observed gains of 1.02 points on SentiCoref (interval includes zero) and 1.71 on NewsMTSC, at substantially different inference costs.
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2611.02989 [cs.CL] |
| (or arXiv:2611.02989v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2611.02989 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Nishan Chatterjee [view email]
[v1]
Mon, 2 Nov 2026 11:59:44 UTC (6,749 KB)
来源：arXiv cs.CL · arxiv.org