ESG-Graph: Hierarchical Residual Graph Attention Network with Analyst-Defined ESG Taxonomy
ESG-Graph: 階層的残差グラフ注意ネットワークとアナリスト定義ESGタクソノミー (AI 翻訳)
Yasser Elouargui, Abdellatif Sassioui, M. Chergui, Rachid Benouini, Mohamed Elkamili, E. Benyoussef, Mohammed Ouzzif
🤖 gxceed AI 要約
日本語
本論文は、ESGテキスト分類のための軽量かつ解釈可能なグラフベースフレームワークESG-Graphを提案する。ESRSベースのタクソノミーを利用し、グラフ注意ネットワークにより効率的な分類を実現。実験では、トランスフォーマーモデルと同等の性能を保ちながら、最大60倍の省エネを達成した。また、政策との整合性や解釈可能性の向上も示されている。
English
This paper introduces ESG-Graph, a lightweight and interpretable graph-based framework for ESG text classification. It leverages a taxonomy based on the European Sustainability Reporting Standards (ESRS) and uses a multi-layer Graph Attention Network. Experiments show comparable performance to efficient transformers while using up to 60× less energy and 10× fewer parameters. The method demonstrates policy alignment and interpretability.
Unofficial AI-generated summary based on the public title and abstract. Not an official translation.
📝 gxceed 編集解説 — Why this matters
日本のGX文脈において
日本ではSSBJによる開示基準が策定中であり、欧州ESRSに基づく本手法は参考になる。また、計算資源の省エネ化は日本のGX推進にも貢献する。
In the global GX context
This work tackles the energy consumption issue in ESG NLP, achieving up to 60× energy savings while maintaining performance. It also shows how regulatory taxonomies like ESRS can be integrated into graph models to enhance interpretability and policy alignment, which is relevant for global disclosure frameworks such as ISSB.
👥 読者別の含意
🔬研究者:This paper provides a novel graph-based approach for ESG text classification that is both efficient and interpretable, offering a promising direction for future research in green AI for sustainable finance.
🏢実務担当者:The ESG-Graph framework can be adopted by companies to classify ESG disclosures with lower computational cost, supporting sustainability reporting efforts.
🏛政策担当者:The integration of ESRS taxonomy into the model demonstrates how regulatory standards can be operationalized, which could inform the design of disclosure frameworks.
📄 Abstract（原文）
Environmental, Social, and Governance (ESG) text classification is important for applications in sustainable finance. However, it remains a challenging task due to domain terminology and regulatory constraints. While transformer-based models achieve strong predictive performance, they often lead to high energy costs and provide limited interpretability. To address these limitations, we introduce ESG-Graph, a lightweight and interpretable graph-based framework for modeling ESG disclosures. In our approach, each sentence is represented as a token-level dependency graph augmented with virtual nodes initialized from a European Sustainability Reporting Standards (ESRS)-based taxonomy, enabling the addition of new ESG concepts without retraining. A multi-layer Graph Attention Network is used instead of transformer encoders, allowing grammatical structure and domain semantics to be modeled jointly. Experiments on three ESG benchmark datasets show that ESG-Graph achieves performance comparable to efficient transformer baselines while consuming up to 60× less energy and using 10× fewer parameters. Additional attribution and ablation studies suggest the method’s policy alignment, interpretability, and robustness.
🔗 Provenance — このレコードを発見したソース
- semanticscholar https://doi.org/10.3390/technologies14050258first seen 2026-05-15 18:38:47
🔔 こうした論文の新着を逃したくない方は キーワードアラート に登録（無料・3キーワードまで）。
gxceed は公開メタデータに基づく研究支援データセットです。要約・翻訳・解説は AI 支援で生成されています。 最終的な解釈・検証は利用者が原典資料に基づいて行うことを前提とします。