metaGraphos: a Web-based system for transcribing, proofreading and publishing scanned documents

Abstract

Purpose

This study aims to present metaGraphos, a crowdsourcing system that aids in the transcription and semantic enhancement of scanned documents by using a pool of volunteers or people willing to participate in exchange for a financial reward.

Design/methodology/approach

The metaGraphos can be used in circumstances where optical character recognition fails to produce satisfactory results, semantic tagging or assigning thematic headings to texts is considered necessary or even when ground-truth data has to be collected in raw form.

Findings

The system automatically provides a Web-based interface comprising a static HTML page and JavaScript code that displays the scanned images of the document, coupled with the corresponding incomplete texts side by side, allowing users to correct or complete the texts in parallel.

Social implications

By assisting the parallel transcription and the semantic enhancement of difficult scanned documents, the system further reveals the hidden cultural wealth and aids in knowledge dissemination, a fact that contributes significantly to the academic-scientific dialog and feedback.

Originality/value

Individual researchers, libraries and organizations in general may benefit from the system because it is cost-effective, practical and simple to set up client–server architecture that provides a reliable way to transcribe texts or revise transcriptions on a large scale.

Citation

Varthis, E. and Poulos, M. (2023), "metaGraphos: a Web-based system for transcribing, proofreading and publishing scanned documents", Collection and Curation, Vol. 42 No. 4, pp. 101-110. https://doi.org/10.1108/CC-01-2023-0002