arXiv cs.CL· Artem Suslov, Takehiro Hashimoto·· 12 小时前AI 评分22
将 LLM 标注的文学情感蒸馏进序数 BERT 模型
Distilling Literary Sentiment from LLM Annotations into an Ordinal BERT Model
AI 导读
研究用 Gemma 4 31B 按十点情感量表标注的 10,889 条文学句子，以序数回归微调五个 BERT 模型，并在海明威《老人与海》的 1,923 条人工标注句子上评估。
正文 · 原文
Abstract:The scarcity of human-annotated literary corpora limits the development of domain-specific sentiment analysis models. This study investigates whether LLM-generated annotations can supervise smaller models for literary sentiment analysis. We fine-tune five BERT models using ordinal regression on 10,889 literary sentences annotated by Gemma 4 31B on a ten-point sentiment scale. The models are evaluated against 1,923 human-annotated sentences from Hemingway's "The Old Man and the Sea", alongside the teacher model and conventional baselines.
BERT-large achieves a Kendall's $\tau$ of 0.589, approaching the teacher's 0.604. Although a constant-neutral predictor achieves the lowest overall mean absolute error (MAE), BERT-base outperforms the teacher on non-neutral sentences, achieving an MAE of 0.726 versus 1.116. These findings demonstrate the potential of LLM-generated supervision for developing smaller literary sentiment models, while highlighting the limitations of aggregate error metrics on predominantly neutral datasets.
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2611.02985 [cs.CL] |
| (or arXiv:2611.02985v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2611.02985 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Artem Suslov [view email]
[v1]
Mon, 2 Nov 2026 11:56:12 UTC (45 KB)
来源：arXiv cs.CL · arxiv.org