Introducing the case of digitised historical newspapers
The aim of the CLARIAH-VL Open Humanities Service Infrastructure is to advance digitally-enabled research in Humanities and the Arts by, among other things, providing data-level access to digitised and born-digital resources. In this blogpost series, we will communicate on research scenarios leading to and building upon the datasets made available through CLARIAH-VL. This blogpost introduces the research scenario ‘Newspapers as Data‘.
Introduction
KBR, Royal Library of Belgium’s digitised historical newspaper platform BelgicaPress provides access to over 4.1 million pages of searchable full-text, from over 135 Belgian major newspaper titles published between 1814-1987. These collections have significant potential for digital humanities research. However, platforms such as BelgicaPress are less than ideal for researchers who are looking to build datasets around specific research questions.
Inspired by the ‘Collections as Data’ movement (Padilla et al., 2019; Candela et al., 2020, Ames, 2021 and Candela et al., 2023) as an approach for cultural heritage institutions to prepare their digital collections for analysis using digital methods, CLARIAH-VL has been exploring how to provide data-level access to KBR’s digitised and born-digital collections for digital humanities research. ‘Collections as Data’, enable digital cultural heritage collections to be made available as FAIR (Findable, Accessible, Interoperable and Reusable) datasets.
Dataset
Data-level access to digitised historical newspaper collections, or ‘Newspapers as Data’ has been the focus of a number of ‘Collections as Data’ initiatives, for example at the National Library of Luxembourg, the Digital Library of the Caribbean and CLARIN’s newspaper corpora.
Data-level access means providing access to the underlying files of digitised cultural heritage resources, enabling a fine-grained level of access which will facilitate data analysis by means of tools and methods developed in the field of digital humanities.
This could include offering access to the METS (Metadata Encoding and Transmission Standard) and ALTO (Analysed Layout and Text Object) files (e.g. in XML or JSON); PDFs of the scanned images (e.g. by newspaper issue, or page); and image files, both as JPEG lower-resolution and high-resolution images in TIFF (Tagged Image File Format).
Research Questions
From a Library and Information Science (LIS) perspective, ‘Collections as Data’ marks a sea change in how cultural heritage institutions provide access to their digital collections for humanities researchers. This paradigm shift is accompanied by a number of LIS research questions, such as: How can the data files of digitised historical newspapers be sustainably extracted from collection management systems? How can we provide best guidelines for curating sample datasets for researchers to tempt them to curate FAIR datasets? How can we measure whether the quality of the OCR (Optical Character Recognition) is sufficient for digital humanities research?
Challenges
To facilitate this requirements building process, the CLARIAH-VL project team made use of the “Collections as Data Checklist”, which is currently being developed by the International GLAM Labs Community. This emerging checklist, which was introduced during the Towards implementing Collections as Data in GLAM institutions webinar in October 2022, is intended to provide “an easy to apply method to encourage, especially small and medium-size cultural heritage institutions, to publish their digital collections as ‘Collections as Data’”. Additionally, the DATA-KBR-BE project team at KBR, Royal Library of Belgium, used the checklist as a framework to help structure the development of the functional and technical requirements for the DATA-KBR-BE platform.
Solutions
Building on the initial webinar in 2022, the CLARIAH-VL team contributed to the International GLAM Labs Community to develop the “Collections as Data Checklist”. This checklist, is intended to provide “an easy to apply method to encourage, especially small and medium-size cultural heritage institutions, to publish their digital collections as ‘Collections as Data’ (Candela et al., 2023). To increase its usability and robustness, within the context of the common European Data Space for Cultural Heritage, the checklist has been transformed into a reproducible research workflow within the Social Science and Humanities Open Marketplace. A workflow is a sequence of steps describing how to perform a task within the research data lifecycle (see: How to create a workflow in the SSH Open Marketplace?).
Next Steps
Within the context of CLARIAH-VL, SIC 2: Aggregate is collaborating with the DATA-KBR-BE project team to develop Newspapers as Data. The resulting curated datasets will be published in the Social Sciences and Digital Humanities Archive (SODHA). Where best to publish the humanities research datasets resulting from CLARIAH-VL is currently being explored. Possibilities include the CLARIAH-VL Zenodo Community.
References
Ames, Sarah. 2021. “Transparency, Provenance and Collections as Data: The National Library of Scotland’s Data Foundry.” LIBER Quarterly 31 (1): 1–13. https://doi.org/10.18352/lq.10371
Candela, G., Sáez, M. D., Escobar Esteban, Mp., & Marco-Such, M. (2020). Reusing digital collections from GLAM institutions. Journal of Information Science. https://doi.org/10.1177/0165551520950246
Candela, G., Gabriëls, N., Chambers, S., Pham T-A., Ames, S., Fitzgerald, N., Hofmann, K., Harbo, V., Potter, A., Ferriter, M., Manchester, E., Irollo, A., Van Keer, E., Mahey, M., Holowinia, O and Dobreva, M. (2023) A Checklist to Publish Collections as Data in GLAM Institutions. https://doi.org/10.48550/arXiv.2304.02603
Padilla, T., Allen, L., Frost, H., Potvin, S., Russey Roke, E. & Varner, S. (2019). Final Report : Always Already Computational: Collections as Data. http://doi.org/10.5281/zenodo.3152935 & https://osf.io/mx6uk/wiki/home/
