Research Data in the Humanities: Present, Preserve and Share

Digital resources in the humanities – text, images, audio and video files – show large variation in terms of quality and provenance, while at the same time constantly increasing in number. As a consequence, requirements for digital (and collaborative) editing, annotating, and modelling of data are subject to constant change. Digital editions, for instance, no longer just offer the potential to annotate text and images through established standards (XML/TEI), but also to retrieve entities in an interoperable format (RDF) and to link them to other digital objects (LOD) in other editions, databases, and archives. The digital annotation of both text and/or image data can be published as a digital object itself. Thus, the relation between digital objects turns into a research object in itself, which can be explored, analysed, and searched by a given set of tools and technologies.

A fundamental challenge, however, is that a vast – and still continuously growing – amount of digitised sources is not searchable via one interface or entry point. Search results of numerous different archives, libraries, or other databases cannot be compared automatically; and search requests – as they have to be performed in a de-centralised manner – cannot be (centrally) saved and re-used. Established generic models, together with the necessary expertise in dealing with individual project data as digital objects, are still lacking practice and broad usage in digital humanities and cultural studies. This applies for both data modelling as well as for long-term archiving and persistent publishing of primary digital data.

The proposed poster will illustrate these core aspects of generating, modelling, archiving, exploring, and presenting humanistic research data from a structural and methodological perspective. The digital infrastructure DARIAH-DE is developing a repository for humanities’ research data ( https://de.dariah.eu/repository) to be archived and published in a sustainable and persistent way, allowing for re-use, citation, preservation, and further exploration:

• Persistent Identifiers, e.g. Digital Object Identifiers (DOI), enable scholarly users to identify research data distinctly. They serve as a crucial pre-condition for archiving, retrieving and citing primary digital data not only in science, but in the humanities as well.
• The modelling of digital objects with established standards (CIDOC, SKOS etc.) and formats (RDF) provides the basis for generating links, thus allowing for generating new knowledge on objects, collections, and/or corpora. Modelling data with the RDF standard allows for linking different objects and therefore accessing information, which can be encoded in variable formats and stored in different databases and archives. The poster will show some practical examples for generic search across TEI-encoded corpora.
• Digital data collections can serve as innovative sources for humanities’ research. As such, they can potentially generate new knowledge. Which are mandatory prerequisites for dealing adequately with information in image- and object- orientated research? Which standards for structuring research data apply, enabling a comprehensive search functionality also across heterogeneously structured data collections and archives? The poster will also present the DARIAH-DE Data Federation Architecture, a set of different tools and services creating innovative options of data management, e.g. accessing and linking heterogeneous data sources of various provenances, analysing existing distributed data collections as well as services to generate data and schema interoperability.
• The poster will show how repositories can help users of varying research disciplines to model, store, and publish their data in a standardised, interoperable way, allowing for re-use, sharing, and further exploration.

The poster will accompany a hands-on workshop adressing these questions in more detail, together with (potential) users. DARIAH-DE will present its repository, its features, the Data Federation Architecture and other tools and services related to dealing with and modelling of humantistic research data via demonstrations and hands-on sessions, enabling participants to actively explore the services also with their own data.