Data Science in Libraries: Library Profile

--- tags: liber-dslib --- # Data Science in Libraries: Library Profile We suggest the members of the group fill it out for their libraries as a start, probably no need for a survey. Or this can be seen as a kind of small collaborative survey. ! The whole matrix is completed for every library. If any of the four data science aspects (Collections as Data, Library Support, ...) are not relevant for a specific library, it can be left empty. This will help to acquire a more generic overview of data science in libraries; also the differences across libraries will become more visible. Responses from the matrix can be summarised as separate paragraphs for the report. And we could have more elaborate cases picked from this for the final report. | | Collections as Data | Library Support | Research Support | Research Intelligence | | ------------------------------| --------------------|-----------------| ------------------| ----------------------| | Current activities | | | | | | Organisation | | | | | | Funding | | | | | | Staff & skills | | | | | | Partners | | | | | | Ethics & Legal aspects | | | | | | Plans for future | | | | | 1. Current activities: Describe the current data science activities undertaken in the library. In what services is data science used? What data science aspects? 2. Organisation: Describe the organisation of data science activities within the library. Is the data science work organised in a specific team? Is it carried out by one or several individual staff members? Or is data science perceived more as a skillset that is integrated in the different library services and used by different staff members? Is the organisation more top-down or bottom-up? 3. Funding: How are data science activities funded? Is there permanent or temporary funding available? Is it project-based or otherwise? 4. Staff & Skills: Which library staff members are involved in data science within library? From which departments? Are external data scientists involved? How? What training opportunities are available? 5. Partners: With whom does the library collaborate in relation to data science activities? Who are the main partners? How are they involved? 6. Ethics & legal aspects: How are ethics & legal aspects in relation to data science addressed within the library? 7. Plans for future: What library data science activities are planned in the future? Note: Topics 2, 3, and 4 could be merged into one 'Organisation'; If needed, topic 6 can be split into two: 'Ethics' & 'Legal aspects'. To do by WG meeting in January: 1. try out the matrix for University of Antwerp Library and the National Library of Sweden; 2. extend the description of four data science activities types; different people involved depending on their expertise and interests. *** *Matrix for the National Library of Sweden (NLS)* | | Collections as Data | Library Support | Research Support | Research Intelligence | | ------------------------------| --------------------|-----------------| ------------------| ----------------------| | Current activities |Digitization as well as preservation of digital material of NLS collections is on-going. KB-labb (start 2019) prepare methods for data analysis (eg Swedish BERT model). ||Some support for researchers working with the KB-labb.|Research intelligence for OA publications and national consortium negotiations.| | Organisation |Digitization is a core activity of NLS. The work is organised by one unit within NLS. This unit collaborate within and outside the NLS. KB-labb was until recently a project initiated by NLS but will soon become permanent.| |The support for researchers regarding their data analysis is organised within KB-labb.|This work is organised within the Unit for Research Collaboration.| | Funding |NLS budget.| |NLS budget + researchers' project budgets.|NLS budget.| | Staff & skills |Data scientists in KB-labb.| |Data scientists in KB-labb.| | | Partners |Research libraries w large collections. Riksarkivet (National Archive).| | | | | Ethics & Legal aspects |unclear.| |unclear.|unclear.| | Plans for future |TBA | |TBA |TBA | *** *Matrix for University of Antwerp Library* | | Collections as Data | Library Support | Research Support | Research Intelligence | | ------------------------------| --------------------|-----------------| ------------------| ----------------------| | Current activities |The library hosts several heritage collections (university archive and collections of historical manuscripts and works of art). Since 2011, the library is gradually digitising these collections: https://www.uantwerpen.be/en/projects/archives-heritage-library-university-antwerp/special-collections-university-antwerp/digital/digital-platform/. At the moment, the library automation team is implementing IIIF framework that would allow a more flexible interaction with images within the collection. is AAT also part of this? https://www.getty.edu/research/tools/vocabularies/aat/|The different departments within the library regularly produce reports and statistics both for internal use and for external organisations (primarily, regional or national). At the moment, this work is carried out using a combination of tools within the library system and MS Excel. Some staff members export data from the library system as csv and then continue with R or Python. |Library offers a possibility for research to deposit research data. A service is being developed that would allow to retrieve also metadata of data sets that have been deposited in external repositories (e.g. Zenodo). Ocassionaly, library receives a request from researchers to provide data for research. This can be a bibliographic metadata collection, images, or full-texts that are available in the library system. | Research intelligence activities are primarily carried out by the Department of Research Affairs & Innovation (DRAI, not part of the library). The library staff, however, provide data from the institutional repository to DRAI. In 2021, two departments within the library collaborated and produced a report on Open Access publications for the Faculty of Arts. | | Organisation |This is done by staff in 'Special Collections' department of the library. The development of IIIF is by library automation team.|Different library departments: metadata & acquisitions, electronic resources, readers' services, institutional repository, etc. These activities are part of regular work routines. | Staff of the institutional repository. | Done by an external department. Library only provides data. | | Funding | ?| Library budget| Library budget| - | | Staff & skills |3 people (?) at the special collections' department. No special training in DS.|Heads and individual members of the different library departments. Self-taught in using MS Excel, Python, SQL, Jupyter Notebooks. However, the level of skills is very varied.|2 people at the institutional repository department. |-| | Partners |?| ?|Department of Research Affairs & Innovation, ? |Department of Research Affairs & Innovation, ?| | Ethics & Legal aspects |?|?|Some of statistics are required by national and regional legal frameworks.|Some of statistics are required by national and regional legal frameworks.| | Plans for future |?|In December 2021, the library launched a 2-year project the goal of which to make library data available for computational methods. The focus of the project is on automation of routine analytics carried out by library staff. |?|There is an idea to develop a kind of Open Science monitor using the library data (both bibliographic metadata and metadata of research data), but no work has been started yet.| Notes: - Overlap between collections as data and research support. When writing about requests from researchers to provide data from the library. It was not clear where exactly should this be written: in the collections as data or research support. It would be helpful to clarify what exactly is meant in each section. - What counts as data science activity here? In the section 'collections as data', the only thing I could think of was digitisation, but that's not really a ds activity. It can be seen as a kind of support for ds activities (e.g. image processing), but not ds in itself. If digitisation is not considered, then most likely this section would be left empty.