Digimälu/Digital Memory 2021 | Eesti Rahvusraamatukogu

Digimälu/Digital Memory 2021

AI – loome targemat kultuuripärandit?
AI – Making Cultural Heritage Smarter?

Digiajastu on kultuuripärandi sektorile juba suuri muudatusi toonud, kuid muutused pole veel lõppenud. Tehisintellekti ja masinõppe kasutamine kultuuripärandiasutuste poolt on kasvamas, luues uusi võimalusi kultuuripärandi kasutamiseks.

Eesti Rahvusraamatukogu korraldab ühepäevase tasuta seminari, et tutvustada uuenduslikke projekte, mis kasutavad tehisintellekti kultuuripärandi hüvanguks. Arutame kultuuripärandi tulevikupotentsiaali üle ning uurime, mida tehisintellektil on kultuuripärandile ja mäluasutuste digitaalsele tulevikule tervikuna pakkuda? Veebiseminar toimub inglise keeles. 

The digital age has already transformed the cultural heritage sector, yet the transformation has not ended. The use of artificial intelligence and machine learning by cultural heritage institutions is increasing, creating new ways of using cultural heritage.

The National Library of Estonia organises a free full-day webinar to introduce innovative projects using AI for the benefit of cultural heritage. The webinar is held in English and looks into the future potential that AI could offer for cultural heritage and the digital future of memory institutions as a whole.

Registration: https://forms.office.com/r/mUhTkyX6J5

Time: October 21, 2021 at 11.00 EEST
Live Broadcast: Sli.do webpage
Contact: Laura Annast, Laura.Annast@nlib.ee, Communication Specialist
Organizer: National Library of Estonia and eodopen.eu

SCHEDULE

11.00—11.10 Opening words
Moderator Raivo Ruusalepp, Manager, EY Consulting

11.15—11.45 How Europe’s digital future shapes memory institutions
Ott Velsberg, Government Chief Data Officer of Estonia

11.50—12.20 AI and Copyright Exceptions – an international comparison
Ben White, Researcher at the Centre for Intellectual Property Policy & Management, Bournemouth University

12.25—12.55 Fairness in AI Empowered Access and Exchange
Helen Eenmaa, Researcher in Information Technology Law, University of Tartu School of Law 

13.00—13.30 Experimenting with AI/ML at the National Library of Estonia
Jane Makke, Head of Library Services Development Centre, National Library of Estonia

14.00—14.25 Using AI to enrich cultural images: the Saint George on a Bike project
Eleftheria Tsoupra, Technical Analyst (Europeana Foundation), R&D team

14.30—14.55 Aiming at getting the most out of museum collection data
Mirjam Rääbis, Estonian National Heritage Board
Tiit Sepp, STACC OÜ

15.00—15.25 Historical Newspaper Content Mining: insights from the impresso project
Maud Ehrmann, Research scientist at EPFL (École polytechnique fédérale de Lausanne) Digital Humanities Laboratory

15.30—15.55 Image Annotations for Visually Impaired – Assisted Hybrid Workflows
Dr. Christoph Lofi, Assistant Professor TU Delft, Netherlands

16.00—16.25 How AI tools can help overcome information overload
Ruth Pickering, Co-Founder and Chief Operating Officer at Yewno

PRESENTATIONS

Ott Velsberg, Government Chief Data Officer of Estonia



What will Europe's digital future look like and what impact will it have on Estonia? How do European digital and data strategies relate to the Digital Agenda 2030 for Estonia? What is the role of memory institutions in the digital society in the next decade?

Ben White, Researcher at the Centre for Intellectual Property Policy & Management, Bournemouth University

Machine learning is predicated on the analysis of data – often “big data”. This in the case of cultural heritage organisations is often sourced from third parties and therefore subject to intellectual property rights, or if digitised from the public domain will often attract sui generis database rights in Europe. Given this, copyright exceptions to allow lawful data analysis of materials are a prerequisite. This presentation will discuss the new text and data mining exceptions in the Copyright in the Digital Single Market Directive as well as compare them to similar provisions in the US and Japan.

Helen Eenmaa, Researcher in Information Technology Law, University of Tartu School of Law 

When asking about justice and ethics concerning AI empowered digital memory, discussions could range from the ramifications of societies’ legal choices and potential ethical implications of technologies we use to the fairness we owe to future generations. This talk will combine the big picture with examples on the ground. In particular, it discusses the work done in reCreating Europe consortium on digital copyright law and the use of technologies for the democratization of cultural practices.

Jane Makke, Head of Library Services Development Centre, National Library of Estonia

In a desire to digitalize the daily workflows, improve user experience and enrich the data, the National Library of Estonia has recently started the experiments with AI and ML. This presentation will allow us to take a peek on these efforts, learnings and conclusions, as well as on the plans for the future.

Eleftheria Tsoupra, Technical Analyst (Europeana Foundation), R&D team

To enable access to collections of cultural heritage images, platforms like Europeana require rich metadata. This presentation will introduce Saint George on a Bike (SGoaB), a joint project between the Barcelona Supercomputing Center and Europeana Foundation, whose goal is to enrich the metadata associated with European cultural heritage images, by using AI content analysis and machine learning tools. The progress that the project has made so far as well as the challenges faced will be shortly presented during this session. 

Mirjam Rääbis, Estonian National Heritage Board and Tiit Sepp, STACC OÜ

 

Museums have used a Centralised Museum Information System (MuIS) in Estonia for more than ten years. The system contains over 4,6 million museum objects with technical and historical descriptions available for usage. This naturally has increased the possibilities and intentions of using the data to improve the service, data analysis, AI development, engaging the users, etc. We will take a closer look at these ambitions vs. the reality of how the data can actually be utilized. We will focus on a project named “Sälli”, a prototype that uses machine learning to assess museum objects' preservation.

Maud Ehrmann, Research scientist at EPFL (École polytechnique fédérale de Lausanne) Digital Humanities Laboratory

Media Monitoring of the Past* is an interdisciplinary research project in which a team of computational linguists, designers and historians collaborate on the datafication of a multilingual corpus of digitized historical newspapers. The primary goals of the project are to improve text mining tools for historical text, to enrich historical newspapers with automatically generated data, and to integrate such data into historical research workflows by means of a newly developed user interface**. Beyond the challenges specific to the different research areas underpinning each of these objectives, the question of how best to adapt text mining tools and their use by humanities scholars is at the heart of the impresso enterprise. In this presentation we will discuss our efforts to overcome the challenges posed by large collections of digitised newspapers and to integrate text mining and data visualisation applications in general historical research practices.
* https://impresso-project.ch
** https://impresso-project.ch/app

Dr. Christoph Lofi, Assistant Professor TU Delft, Netherlands

Annotating images in various publication types are becoming relevant and important, especially with visually impaired users in mind. However, the process of creating such image annotations usually relies on expensive experts, thus limiting the scale of the process. Furthermore, experts report that creating such annotations from scratch is mentally taxing due to high cognitive load. The goal of our work is to assist the manual annotation process with automatic tools and heuristics, reducing the required cognitive load and improving the efficiency of the overall process.

Ruth Pickering, Co-Founder and Chief Operating Officer at Yewno

In this presentation we will review how AI can improve indexing and add new capabilities such as context and visualisation. This can help with historic barriers to successful research such as fragmented information and poor tagging. It can also help with historic bias where terms have been missed, and the evolution of language.

Book recommendations

A selection of books from the National Library’s collections on AI, machine learning and digital technology