A Method for Automatic Verification and Correction of Document Formatting in Educational Contexts

A Method for Automatic Verification and Correction of Document Formatting in Educational Contexts

Akkuzina V.M. (NRU HSE, Perm, Russia)
Vikentyeva O.L. (NRU HSE, Perm, Russia)

Abstract

This paper proposes a method for the automatic verification and correction of formatting in educational and research documents, designed to support the diverse local formatting standards used by academic institutions. Existing tools are either rigid rule-based systems bound to a single institutional template, incapable of automatic correction, or machine learning approaches that analyze document structure without providing deterministic, rule-based correction. To close this gap, the proposed method combines three components: a tree-structured, extensible template that formalizes local formatting rules for each paragraph type; a classification algorithm that identifies paragraph types from a combination of visual, structural, and contextual features; and a correction mechanism that maps detected inconsistencies to concrete corrective actions. The method was implemented as a client-server prototype built on ASP.NET Core and the OpenXML library, with the formatting template stored as structured XML. The prototype was evaluated on a sample of 10 student reports (1,018 paragraphs) previously reviewed by teaching assistants at HSE University – Perm. Paragraph classification reached an accuracy of 87.2%, with most confusions arising between lists, plain text, and headings in documents lacking style-based formatting. Formatting inconsistencies were detected with a recall of 90.8%, and 83.1% of the identified violations were resolved automatically without introducing new inconsistencies. Automatic verification of a 15–25-page document took about one minute, compared with the 15–30 minutes reported by most teaching assistants for manual review, reducing verification time by more than a factor of ten. These results demonstrate that combining an adaptable rule template with multi-feature paragraph classification enables accurate, transparent, and reusable formatting verification across different institutional standards, while automatic correction remains limited to locally resolvable inconsistencies.

Keywords

automatic formatting verification; formatting correction; paragraph classification; formatting template; educational documents; formatting automation.

Edition

Proceedings of the Institute for System Programming, vol. 38, issue 5, 2026, pp. 325-336

ISSN 2220-6426 (Online), ISSN 2079-8156 (Print).

DOI: 10.15514/ISPRAS-2026-38(5)-18

For citation

Akkuzina V.M., Vikentyeva O.L. A Method for Automatic Verification and Correction of Document Formatting in Educational Contexts. Proceedings of the Institute for System Programming, vol. 38, issue 5, 2026, pp. 325-336 DOI: 10.15514/ISPRAS-2026-38(5)-18.

Full text of the paper in pdf (in Russian) Back to the contents of the volume