Environmental, Social and Governance Sentiment Analysis on Slovene News: A Novel Dataset and Models

TL;DR

This paper presents the first Slovene ESG sentiment dataset and models for automatic detection, evaluating various models including LLMs and fine-tuned BERT variants, with promising results for environmental, social, and governance aspects.

Contribution

Introduces a novel Slovene ESG sentiment dataset and compares multiple models, including LLMs and fine-tuned transformers, for ESG sentiment analysis.

Findings

LLMs outperform other models on Environmental and Social aspects.

Fine-tuned SloBERTa performs best on Governance classification.

The best model enables case studies on ESG trends over time.

Abstract

Environmental, Social, and Governance (ESG) considerations are increasingly integral to assessing corporate performance, reputation, and long-term sustainability. Yet, reliable ESG ratings remain limited for smaller companies and emerging markets. We introduce the first publicly available Slovene ESG sentiment dataset and a suite of models for automatic ESG sentiment detection. The dataset, derived from the MaCoCu Slovene news collection, combines large language model (LLM)-assisted filtering with human annotation of company-related ESG content. We evaluate the performance of monolingual (SloBERTa) and multilingual (XLM-R) models, embedding-based classifiers (TabPFN), hierarchical ensemble architectures, and large language models. Results show that LLMs achieve the strongest performance on Environmental (Gemma3-27B, F1-macro: 0.61) and Social aspects (gpt-oss 20B, F1-macro: 0.45), while fine-tuned SloBERTa is the best model on Governance classification (F1-macro: 0.54). We then show in a small case study how the best-preforming classifier (gpt-oss) can be applied to investigate ESG aspects for selected companies across a long time frame.