---
title: "超越情感：从金融新闻中提取结构化信息的新框架"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2607.28496v1"
permalink: "https://feed.astralor.com/items/itm_a22c7e313fed28868e23e029"
published: "2026-07-30T16:41:55.000Z"
score: 73
channel: "finance"
tags: ["FinBERT","FNSPID","LLaMA-3.1-70B","结构化信息提取","金融NLP"]
---

# 超越情感：从金融新闻中提取结构化信息的新框架

## AI 点评

> 这篇论文的价值在于用最高效的方式证明了'情感分数不够用'这个老直觉是对的，并且给出了可复现的框架和清晰的量化证据。对做金融舆情或股价预测的读者来说，文中53.5%的分歧率和组合后0.600的F1是值得直接借鉴的基线。不过要留意，LLM提取结构化特征的成本远高于FinBERT，实际应用中得权衡性价比。

## 速览

金融情感分析虽已成为新闻驱动股价预测的标准工具，但仅将多维新闻压缩为单一极性分数。研究者提出基于LLaMA-3.1-70B的结构化信息提取框架，从FNSPID数据集4.2万条新闻-股票对中提取事件类型、影响范围、时间跨度等六个语义维度。实验发现FinBERT情感特征在非线性模型下预测力较强（F1=0.576），但在线性模型下较弱（F1=0.230）；LLM提取的结构化特征与情感特征存在53.5%的系统性分歧，两者结合显著提升至F1=0.600，优于任一单独信号。该研究表明情感与语义解耦在金融文本中是系统性的且可利用的，为多维金融NLP开辟了新方向。

## 关键点

- 提出基于LLaMA-3.1-70B的结构化信息提取框架，从金融新闻中提取六个语义维度，超越单一情感分数。
- 在41,618个新闻-股票对上实验，发现FinBERT情感特征在非线性模型下表现优异（F1=0.576），线性模型下仅0.230，揭示高度非线性的情感-收益关系。
- 结构化特征与情感特征存在53.5%的系统性分歧，两者结合获得F1=0.600，显著优于任一单独信号（p<0.0001），并在全部七种事件类型上一致提升。
- 特征重要性分析显示六个维度贡献均衡（14%-21%），证明压缩新闻为单一情感分数会导致大量信息损失。

## 全文

Financial sentiment analysis has become a standard component in news-driven stock prediction, yet it reduces rich, multi-dimensional news articles to a single polarity score. We hypothesize that financial news encodes multiple orthogonal information dimensions---event type, impact scope, temporal horizon, and semantic confidence---that sentiment alone cannot capture, and that these dimensions carry independent predictive value. To test this hypothesis, we propose a structured information extraction framework that leverages LLaMA-3.1-70B to extract six semantic dimensions from financial news. Through large-scale experiments on 41,618 news--stock pairs from the FNSPID dataset, we find that (i) FinBERT sentiment features exhibit strong predictive power under nonlinear models (F1=0.576) but substantially weaker performance under linear models (F1=0.230), revealing a highly nonlinear sentiment--return relationship; (ii) LLM-extracted structured features, while individually weaker, capture information orthogonal to sentiment, as evidenced by a 53.5% systematic disagreement rate between the two approaches; and (iii) combining both signal sources yields F1=0.600, significantly outperforming either alone ($p < 0.0001$), with consistent improvements across all seven event types. Ablation experiments confirm that non-sentiment structural dimensions (event type, impact subject, time horizon, confidence) independently contribute $Δ\text{F1} = +0.019$ beyond FinBERT alone. Feature importance analysis reveals balanced contributions from all six extracted dimensions (14--21%), demonstrating that compressing news into a single sentiment score incurs substantial information loss. Our results suggest that the sentiment--semantics decoupling in financial text is systematic and exploitable, opening a new direction for multi-dimensional financial NLP.

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_a22c7e313fed28868e23e029