Data modeling and processing as a library service for digital humanities research data management

Abstract

Our University Library (UBL) have seen the need and potential for strengthening the infrastructure for digital full text resources at the University of Bergen and we wanted better to establish the library’s role in this area.

Five years ago, several research communities expressed concerns that it was becoming increasingly difficult to sustain the competence necessary to run and maintain both physical and digital research archives. More specifically a concrete need for supporting XML-based digital humanities text resources was voiced. We felt the UBL could meet this need by providing a new service.

A combination of data modeling, data conversion and an active use of open data solutions has in our view shown itself to be an effective solution. We find that in-house data modeling and processing competence is essential in order to cope with tasks connected to digital text and image resources.

Our poster will outline our digital service provision by giving selected references and examples.

The Wittgenstein Archives at the University of Bergen (WAB) is one example of a recipient of UBL data services. WAB maintains a richly encoded XML version of the complete Nachlass of philosopher Ludwig Wittgenstein.

A library web resource building upon library data modeling and conversion is MARCUS, which shows how catalogue data and image data for the University Library’s own manuscript collections and photographic collections are currently digitized and interconnected using electronic representations of documents and Linked Data/RDF (Resource Description Format) metadata. MARCUS meets UBLs long felt need for a unified special collections digital system. This relates not only to document storage, display and dissemination, but also to the library workflow for the special collections. Both WAB and MARCUS benefits, strategically and day-to-day, from the same competencies within the library.

We think that a sensible future-oriented solution entails that each institution, to a greater degree than before, works with modeling and conversion of its own data. Our view is that using Linked Data/RDF encoding will pave the way to connecting data sets in such a way that they enrich one another. Rather than functioning as system providers, we envision large institutions processing and sharing open datasets, as well as encouraging and enabling others to do the same.

In line with LIBERs Ten recommendations for libraries to get started with research data management our view is that data modeling and data conversion, within the frame of an active use of open data solutions, are services that belongs within the portfolio of the research library.

Presented by Irene Eikefjord, Senior Librarian, University of Bergen Library