Pharos-ESG：面向 ESG 报告的多模态解析、上下文叙述与层级标注框架
Pharos-ESG: A Framework for Multimodal Parsing, Contextual Narration, and Hierarchical Labeling of ESG Report
AI 导读
arXiv 论文提出 Pharos-ESG 框架，通过基于版面流阅读顺序建模、目录锚点引导的层级切分和多模态聚合，将 ESG 报告转化为结构化表示，并附加 ESG、GRI 与情感标签。在标注基准上的实验显示其性能持续优于专用文档解析系统和通用多模态模型。作者还发布了 Aurora-ESG，首个覆盖中国大陆、香港和美国市场的大规模公开 ESG 报告数据集，含细粒度版面与语义标注。
正文 · 原文
Abstract:Environmental, Social, and Governance (ESG) principles are reshaping the foundations of global financial governance, transforming capital allocation architectures, regulatory frameworks, and systemic risk coordination mechanisms. However, as the core medium for assessing corporate ESG performance, the ESG reports present significant challenges for large-scale understanding, due to chaotic reading order from slide-like irregular layouts and implicit hierarchies arising from lengthy, weakly structured content. To address these challenges, we propose Pharos-ESG, a unified framework that transforms ESG reports into structured representations through multimodal parsing, contextual narration, and hierarchical labeling. It integrates a reading-order modeling module based on layout flow, hierarchy-aware segmentation guided by table-of-contents anchors, and a multi-modal aggregation pipeline that contextually transforms visual elements into coherent natural language. The framework further enriches its outputs with ESG, GRI, and sentiment labels, yielding annotations aligned with the analytical demands of financial research. Extensive experiments on annotated benchmarks demonstrate that Pharos-ESG consistently outperforms both dedicated document parsing systems and general-purpose multimodal models. In addition, we release Aurora-ESG, the first large-scale public dataset of ESG reports, spanning Mainland China, Hong Kong, and U.S. markets, featuring unified structured representations of multi-modal content, enriched with fine-grained layout and semantic annotations to better support ESG integration in financial governance and decision-making.
| Subjects: | Artificial Intelligence (cs.AI) |
| ACM classes: | I.2.7 |
| Cite as: | arXiv:2511.16417 [cs.AI] |
| (or arXiv:2511.16417v4 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2511.16417 arXiv-issued DOI via DataCite |
Submission history
From: Yu Zou [view email]
[v1]
Thu, 20 Nov 2025 14:41:44 UTC (5,894 KB)
[v2]
Wed, 25 Mar 2026 11:17:23 UTC (5,894 KB)
[v3]
Sun, 29 Mar 2026 07:20:09 UTC (4,573 KB)
[v4]
Thu, 29 Oct 2026 14:14:07 UTC (4,573 KB)
来源：arXiv cs.AI · arxiv.org