CF-RAG: A Dataset and Method for Carbon Footprint QA Using Retrieval-Augmented Generation

2025-08-05

AI Summary

Extracting carbon footprint information from non-standardized PDF sustainability reports is challenging due to mixed text–table layouts, heterogeneous formatting, and absence of uniform structure. Method: We propose CF-RAG, the first framework for carbon footprint question answering over unstructured text derived from PDF parsing. It jointly models tabular and textual content, fine-tunes Llama-3 with retrieval-augmented generation (RAG), and enhances multimodal understanding of PDF documents. Contribution/Results: We introduce CarbonPDF-QA—the first fine-grained, human-annotated dataset for this task—and demonstrate that CF-RAG significantly outperforms GPT-4o and state-of-the-art table–text joint reasoning models on carbon footprint QA, achieving substantial accuracy gains. This work establishes a novel paradigm for automated, structured analysis of sustainability reporting.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Question Answering
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Systems and Infrastructure for Web, Mobile and WoT: Sustainability and carbon-aware systems for Web, mobile, and WoT
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work

Abstract

Product sustainability reports provide valuable insights into the environmental impacts of a product and are often distributed in PDF format. These reports often include a combination of tables and text, which complicates their analysis. The lack of standardization and the variability in reporting formats further exacerbate the difficulty of extracting and interpreting relevant information from large volumes of documents. In this paper, we tackle the challenge of answering questions related to carbon footprints within sustainability reports available in PDF format. Unlike previous approaches, our focus is on addressing the difficulties posed by the unstructured and inconsistent nature of text extracted from PDF parsing. To facilitate this analysis, we introduce CarbonPDF-QA, an open-source dataset containing question-answer pairs for 1735 product report documents, along with human-annotated answers. Our analysis shows that GPT-4o struggles to answer questions with data inconsistencies. To address this limitation, we propose CarbonPDF, an LLM-based technique specifically designed to answer carbon footprint questions on such datasets. We develop CarbonPDF by fine-tuning Llama 3 with our training data. Our results show that our technique outperforms current state-of-the-art techniques, including question-answering (QA) systems finetuned on table and text data.

Problem

• Extracting carbon footprint data from unstructured PDF reports
• Handling inconsistent text and tables in sustainability reports
• Improving QA accuracy for carbon footprint questions

Innovation

• Retrieval-Augmented Generation for carbon footprint QA
• Fine-tuned Llama 3 on CarbonPDF-QA dataset
• Handles unstructured PDF text and table data