NotaryPedia Showcases AI-Powered Platform for Unlocking Malta's Notarial Heritage

The NotaryPedia project held its first public showcase on Thursday 30 April 2026 at the Notarial Archives in Valletta, presenting live demonstrations of the AI-assisted knowledge platform being developed to make over 20,000 volumes of historical Maltese notarial manuscripts searchable and accessible.

The event, hosted by Prof. Charlie Abela from the Department of Artificial Intelligence within the Faculty of ICT, brought together institutional stakeholders, heritage professionals, and research collaborators to witness the project's progress first-hand. Prof. Valerie Sollars, Pro-Rector for Strategic Planning and Sustainability, delivered institutional remarks on behalf of the University. Dr Joan Abela, President of the Notarial Archives Foundation, spoke to the significance of the Archives and the collaborative partnership with the University that underpins the project. Representatives from the University's Knowledge Transfer Office and collaborators of the Foundation were also in attendance.

The Notarial Archives hold one of the most extensive documentary collections in the Mediterranean, spanning the fifteenth to the twenty-first century. The manuscripts record the legal, commercial, social, and personal transactions of Maltese life across five centuries, property sales, marriage contracts, wills, business partnerships, and disputes, but remain largely inaccessible due to the specialist palaeographic expertise required to read them. NotaryPedia aims to change this by combining artificial intelligence with expert human review to transcribe, extract, and structure the knowledge contained within these volumes.

The showcase was structured around the project's knowledge pipeline, with each of the project's six Research Support Officers demonstrating a distinct stage of the system on real manuscript data.

Mr Chukwuma Sidney Anih presented the handwritten text recognition system, showing how AI models trained on historical Maltese notarial hands can transcribe manuscript pages automatically.

Ms Vanessa Buhagiar and Mr Nathan Ciantar followed with a combined demonstration of palaeographic review and the semi-automated guardrailing tooling being developed to make expert quality assurance both rigorous and scalable.

Ms Charlene Ellul demonstrated her work on automatic pre-annotation and the development of the Notarypedia catalogue, an important step towards organising manuscript content into a knowledge graph.

Mr Chimauche Njoku showed how the knowledge graph evolves and enriches itself as new manuscripts are processed, learning new entities and predicting connections in incomplete records.

Mr Brendon Curmi concluded the demonstration sequence with the web platform and workflow orchestration system that integrates these capabilities into a working research tool.

A central theme of the showcase was the project's commitment to human–AI collaboration. Prof. Abela emphasised that NotaryPedia is designed not to replace archival expertise but to amplify it: the AI proposes transcriptions and extractions, human experts review and correct them, and those corrections feed back into the system, making it progressively more capable. This virtuous cycle ensures that scholarly rigour is maintained while enabling the processing of manuscripts at a scale that would be impossible through manual effort alone.

The event concluded with a forward-looking session in which Prof. Abela outlined the project's roadmap, including plans to extend the transcription models to a wider range of historical hands and centuries, expand the knowledge graph across a larger corpus, and develop the platform toward sustained operational use by Archive staff and researchers. He also highlighted the project's potential as a replicable model for AI-assisted cultural heritage preservation, noting that the methods and tools being developed in Malta could be adapted to comparable archival collections across Europe and the Mediterranean.

NotaryPedia is a collaborative initiative between the Department of Artificial Intelligence, Faculty of ICT, University of Malta, and the Notarial Archives Foundation, funded by the Ministry for Culture, Lands and Local Government.