Information Extraction from Fiscal Documents using LLMs

Vikram Aggarwal (), Jay Kulkarni, Aakriti Narang, (), Aditi Mascarenhas, Siddarth Raman (), Ajay Shah () and Susan Thomas ()
Additional contact information
Vikram Aggarwal: Google
Aakriti Narang,: xKDR Forum
Aditi Mascarenhas: xKDR Forum
Siddarth Raman: xKDR Forum
Ajay Shah: xKDR Forum
Susan Thomas: xKDR Forum

No 43, Working Papers from xKDR

Abstract: Large Language Models (LLMs) have demonstrated remarkable capabilities in text comprehension, but their ability to process complex, hierarchical tabular data remains underexplored. We present a novel approach to extracting structured data from multi-page government fiscal documents using LLM-based techniques. Applied to large annual fiscal documents from the State of Karnataka in India, our method achieves high accuracy through a multi-stage pipeline that leverages domain knowledge, sequential context, and algorithmic validation. Traditional OCR methods work poorly with errors that are hard to detect. The inherent structure of fiscal tables, with totals at each level of the hierarchy, allows for robust internal validation of the extracted data. We use these hierarchical relationships to create multi-level validation checks. We demonstrate that LLMs can read tables and also process document-specific structural hierarchies, offering a scalable process for converting PDF-based fiscal disclosures into research-ready databases. Our implementation shows promise for broader applications across developing country contexts.

JEL-codes: H6 H7 Y10 (search for similar items in EconPapers)
Pages: 6 pages
Date: 2025-11
New Economics Papers: this item is included in nep-ain, nep-big, nep-cmp and nep-sea
References: View complete reference list from CitEc
Citations:

Downloads: (external link)
https://papers.xkdr.org/papers/2025Kulkarnietal-acm_icaif.pdf First version, 2025 (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:anf:wpaper:43

Access Statistics for this paper

More papers in Working Papers from xKDR
Bibliographic data for series maintained by Ami Dagli ().