Part of UniTor ESG Insights β€” CIKM 2026

This model is part of UniTor ESG Insights, an interactive human-in-the-loop system for traceable sustainability report analysis. The system integrates specialized models for SDG/GRI alignment, disclosure quality assessment, and climate-related analysis. Predictions can be explored alongside their original source paragraphs and PDF pages.


Task Title

Qualitative vs Quantitative Sustainability Text Identification (QQ)

Task Description

Binary classification model for distinguishing qualitative sustainability disclosures from quantitative disclosures.

Dataset Summary

  • Total samples: 707
  • Qualitative: 589
  • Quantitative: 118

Model

  • Base model: bert-base-cased

Performance

  • Accuracy: 92.1% Β± 3.5
  • Macro F1: 85.7% Β± 6.0

Results are reported as Mean Β± SD across 10-fold cross-validation, where each fold was used once as the test set.

Downloads last month
29
Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support

Model tree for sag-uniroma2/QQ_MODEL

Finetuned
(2,958)
this model

Collection including sag-uniroma2/QQ_MODEL