Mini-ELAG 2021: Program

Where: Microsoft Teams

Date: June 30, 2021 14:00 – 16:00 CEST

You can register here !

We got a program for the Mini-ELAG on 30 June 2021 14:00 - 16:00 CEST. Join us on Microsoft Teams and learn what our ELAG presenters have been doing the last months. This Mini-ELAG will be filled with 5 to 10 minute lightning talks and is free for registered participants.


Image processing in the cultural heritage sector

Tan Lu, Vrije Universiteit Brussel (VUB) and Royal Library of Belgium (KBR), Belgium

This talk will briefly touch three aspects namely layout analysis, damage recognition and quality assessment, where the idea is to show that image processing and mathematical modelling may be applied to assist automation and information extraction in the cultural heritage. In particular, for layout analysis I will introduce our work on a new algorithm named ‘Document Segmentation using Probabilistic Homogeneity’ (ref: where we propose a probabilistic local text homogeneity (PLTH) model for a more reliable document layout parsing. The proposed PLTH model is further incorporated in a Bayesian damage recognition framework, which will be introduced in part 2. Lastly, I will also briefly introduce the idea of a unified image quality assessment using deep learning. All these results are obtained under the ADOCHS ( project based on the collaboration between VUB (Vrije Universiteit Brussel) and KBR (The Royal Library of Belgium).

Duration: 10 minutes

Discover: a modular framework for research data in the project eHumanitie

Jaime Penagos, IT Dpt. University Library, Ludwig Maximilian University Munich, Germany

At the University Library LMU Munich, we established a framework named “Discover”, that presents indexed research data to the end user, based on a self-developed extension of the DataCite schema we called “rdUB”. The research data and metadata reside in a Fedora Repository, which was integrated through Apache Camel with Solr and Blacklight.

Duration: 10 minutes

Linked Data Notifications for the Scholarly Web

Patrick Hochstenbach, Ghent University, Belgium

This talk will report our progress in the Mellon project on Scholarly Communication Using the Decentralized Web. This PhD project at Gent University envisions an alternative scholarly communication system that is research-centric, institution enabled and aligned with decentralized web concepts and technologies. In this vision, researchers use a personal domain and associated storage space as their long-term scholarly hub, and the core functions of scholarly communication (registration, certification, awareness, archiving) are fulfilled in a decoupled manner, that is, each function can simultaneously be fulfilled by multiple parties in different ways.

Duration: 5 minutes

Continuous quality assessment for MARC21 catalogues

Péter Király, Gesellschaft für wissenschaftliche Datenverarbeitung Göttingen, Germany

QA Catalogue is an open source quality assessment tool for MARC21 based catalogues. It reads different serialization formats (binary MARC, XML, Alephseq etc.) and analyses different quality aspects, such as the validity of the data elements and their values, completeness, support of functional requirements, classifications and name authorities, the most frequent data elements, the historical aspects of a catalogue. Last summer, together with the colleagues at Gent University Library we have implemented a continuous quality assessment workflow: every Saturday morning the library publishes their whole catalogue as a single compressed file. QA Catalogue service fetches it, and runs assessment, and refreshes the quality dashboard (which is a web application containing a special discovery interface).

Duration: 10 minutes

Pubby: front end for RDF-Data

Tobias Malmsheimer, Stuttgart Media University, German_y_

At Stuttgart Media University multiple research projects create knowledge graphs and other graph-based database structures for different domains. In the context of these projects we have reimplemented the web frontend “Pubby” for RDF data. “Pubby” now uses a modern web framework (Django), is prepared to be combined with ElasticSearch and can easily be adapted and extended for different use cases. Our presentation will focus on the main functions, setup and configuration and an example extension module.

Duration: 5 minutes

Workshop report from the Handschriftenportal

Leander Seige, Leipzig University Library, Germany

A brief report on the current status of the Handschriftenportal, the new decentralized central portal for all medieval manuscripts in German institutions. It will feature the public alpha system as well as several aspects and issues of decentralized image hosting with IIIF, displaying and annotating TEI in Mirador and further questions

Duration: 10 minutes

Unified Access to controlled vocabularies with BARTOC

Jakob Voß, Verbundzentrale des GBV (VZG), Germany

The vocabulary registry BARTOC contains metadata about several thousand library classifications, thesauri, ontologies and other controlled vocabularies. We will show how this information can be used not only to find controlled vocabularies but also to access their contents in a uniform way.

Duration: 5 minutes

Physical access control in the time of Covid-19

Matthew Phillips, Durham University, United Kingdom

To enable social distancing during the pandemic, we had to severely restrict the number of people using our main Bill Bryson Library. Access is controlled using the Telepen Sentry Juno system. We took advantage of the openness of its database to link it to a booking system, allowing controlled and fair access to the building for academic staff and students.

Duration: 5 minutes

Interactive forms with the functional programming language Elm

Olli Lyytinen, Durham University Library, United Kingdom

In order to simplify our multiple different library resource request forms, we consolidated them all into an interactive form which adapts to the user and their choices in the form. This gives the user an intuitive interface to resource requests hiding options and fields irrelevant to them. We built the form using the purely functional web programming language Elm. This offered us much more customisability than using basic form builders available in content management systems available to us.

Duration: 5 minutes

Establishment of a digital curation curriculum in the Czech Republic

Michal Lorenz and Helena Novotná, Masaryk University, Brno, Czech Republic

In this Lightning Talk, the concept of a curriculum for digital curators at the Department of Information and Library Studies of Masaryk University is presented. As the country lacks professionals educated in this area, we have started to educate information studies students to hold the position of digital curator in all types of memory institutions with the support of the National Museum’s project New Phonograph. We are currently striving to certify this specialisation with domestic national memory institutions to ensure higher visibility and employability of graduates.

Duration: 5 minutes

Using Apache Kafka and Kubernetes to process data and run microservices and library applications

Lionel Walter, University of Basel Library, Switzerland

The Basel University Library has released two new web portals this spring : Memobase (, the portal to Switzerland’s audiovisual cultural heritage, and swisscollections (, the gateway to historical and modern collections in Swiss libraries and archives.

They use a common microservices and data infrastructure, based on Apache Kafka and Kubernetes. In this talk we will describe how it works behind the scenes. This talk is a follow-up from ELAG 2019 Talk : Microservices based on the Kafka Event Hub (

Duration: 10 minutes