Logo Lanfrica

<b>End-to-End Digitisation Workflows with Intelligent Document Processing: A Case Study on Archival </b><b>Document Digitisation</b> (Presentation for the Annual Conference of the South African Society of Archivists (SASA), Cape Town, 9 July 2025)

Domaine:

natural language processing

Type de record:

softwarepaper
Créateur:
Nik
Hôte:avatar

This presentation was delivered at the Annual Conference of the South African Society of Archivists (SASA), held from 8–11 July 2025 in Cape Town, South Africa, under the theme Collaborative Archives and Records Management in the Era of Artificial Intelligence.

The session focused on the challenges, opportunities, and ethical considerations of applying Artificial Intelligence (AI) and Intelligent Document Processing (IDP) to digitisation workflows in African archival contexts. Drawing on case studies digitising theses and historical newspapers, the presentation demonstrates how standards-aligned imaging and machine-learning pipelines can improve efficiency, metadata quality, and long-term preservation readiness, while acknowledging risks around bias, automation gaps, and data protection.

Key topics covered include:

  • Digitisation standards (ISO 19264-1, FADGI)
  • Metadata extraction and enrichment
  • OCR/HTR pipelines
  • Risks in AI-driven digitisation (bias, hallucination, POPIA compliance)
  • Implementation models (on-premise, SaaS, hybrid)
  • Recommendations for African archives

This resource is intended as a practical reference for archivists, records managers, and digital preservation practitioners working to modernise collections management through AI and scalable workflows.

Licenses

Similaires