SLSP and the Swiss Data Landscape

Swiss Library Service Platform (SLSP) was founded in 2017 by 15 higher education institutions and libraries in Switzerland.^[1] SLSP aimed to consolidate the datasets of the six existing library consortia in Switzerland into a national platform, operate a centrally managed library system, and offer centrally managed services for libraries. Today, SLSP drives transformation for trustworthy information in Switzerland with around 500 libraries as part of the network, grouped in 31 so-called institution zones in Alma. In addition to traditional academic libraries, this network also includes the special case of patrimonial libraries – for instance, cantonal libraries.

Heritage institutions face different challenges when it comes to publishing open data. Since libraries often hold material such as archival records or images in addition to classic digitised texts, these boundaries are becoming increasingly blurred, and general questions arise around licensing, metadata, and APIs (Roued-Cunliffe, 2020, Chapter 3). Compared with the situation described in another study (Estermann, 2014), the dissemination of this data in Switzerland has made great progress in the last 10 years. In particular, data that is collected in libraries can nowadays be exposed in various ways. Some of the cultural data of Switzerland is exposed via repositories.^[2] Furthermore, there is already a wide range of cultural and research data available in the common discovery platform swisscovery.^[3] This data includes digital objects (over one terabyte) in the cloud library system Alma (Alma Digital), as well as external sources searchable via Primo VE. A large portion of digital objects is preferably stored outside of Alma due to missing long-term storage options and control in Alma Digital.

Over two million records from over 60 institutional repositories, such as e-periodica,^[4] are searchable in swisscovery through harvesting among around 40 million records of the libraries’ holdings in total. Data can also be made available via portals or aggregators, including the Central Discovery Index (CDI), a large index for scholarly material worldwide provided by Clarivate | Ex Libris, and Europeana, for example. Another approach is to recognise that institutions in Switzerland also register their data via DOI agencies such as DataCite.^[5] This registration makes them part of the persistent identifier (PID) graph and offers added value through the linkage of different entities, e.g. creators, institutions and events.

To do justice to this dynamic data landscape with which SLSP is dealing, several projects, such as a joint project called Connectome with Switch and another project with the Swiss National Library are presented below.

The Case of the Connectome Project

Open Data Navigator: Aggregating Swiss Data Sources

Switch operates the national research and education network of Switzerland. It is a member of GÉANT, a pan-European data network,^[6] and collaborates with the European Open Science Cloud (EOSC). Switch is a strategic partner of SLSP and operates many services that SLSP utilises, such as cloud services (e.g. storage and computing), digital identity (edu-ID), and registry (.ch and .li).

A few years ago, Switch launched a project called Connectome preceded by several labs, for instance, one that addressed the handling of research data in social sciences and humanities (Hauf et al., 2021). This project is supported by Swissuniversities^[7]. Connectome’s objective is to ‘promote reuse of open data from heterogeneous areas’ and ‘support [the] development of open data practices’ (Switch, n.d.). To achieve this aim, it harmonises and links metadata through artificial intelligence from various Swiss sources into a knowledge graph. This data can come from administrations, archives, non-governmental organisations, galleries, libraries, museums, and other open data aggregators. Connectome’s goal is for data to be findable, accessible, interoperable, and reusable, following the FAIR principles.

The aggregated data is presented to the public in two forms:

• Open Data Navigator: https://opendatanavigator.switch.ch/
• Switch Open Data API: https://opendatanavigator.switch.ch/api/graphql

Switch developed the Open Data Navigator, which acts as a national open data aggregator. The discovery experience is enhanced by recommendation systems based on artificial intelligence (language models) and semantic relationships among collected resources.

The data can be filtered by facets, e.g. data types, keywords, or providers (Figure 1). There is also the option of displaying a topic map, which provides an additional visual introduction. Further sources are suggested on the right-hand side by using a k-nearest neighbour search for the query terms. The detailed view of a record contains the aggregated metadata and a link directly to the data or a landing page of the provider.^[8]

Connectome Meets Swisscovery: Finding Synergies

Connectome and swisscovery are two platforms that offer Switzerland access to literature and data. The question emerges as to what synergies arise between the two. The following abstract scenarios are conceivable:

• Connectome harvests swisscovery
• swisscovery integrates Connectome
• (artificially intelligent) linkage between swisscovery ⟷ Connectome
• A new portal

A connection with the research platform swisscovery can offer added value but also raises various questions: Which user stories should be considered for this approach? What does the technical integration look like, e.g. through GraphQL? What dependencies are there with Primo VE from Clarivate | Ex Libris?

A project was launched between SLSP and Switch to investigate these questions further as a proof of concept with the following mission statement:

Key synergy is the guidance of swisscovery end-users to relevant resources related to their search criteria in the Swiss research landscape by means of visualisation and providing contextual information.

The project started in April 2024 and will last until the end of 2025, comprising roughly 200 hours of work. The project team consists of data and discovery experts from both institutions. In detail, the experts will try to familiarise ourselves with the end-user community. Building on this, a solution that adds value and is feasible will be specified. This solution will then be implemented as a prototype, evaluated, and documented. Currently, Phase I is underway, and the progress report consists of two areas that will be approached through several workshops.

The first part addresses data structure clarification and explores technical possibilities between the two platforms (Table 1).

The data structures and access points are very different, but a direct or indirect linkage between the records is possible, covering many entities, including DOI (datasets), ORCID (persons), ROR (organisations), URI, and GND (persons and keywords).

As a basis for all the following measures, it is highly important to establish a joint user community. SLSP has, independently of this project, already developed personas in an Agile project management workshop facilitated by a lecturer in digital business management. These personas were further shaped through collective discussions and insights. Two of them demonstrate a strong connection to research-related activities, which is the expected target group of Connectome. Furthermore, personas that do not use swisscovery are crucial for understanding motivations and research conduct locations (Figure 2). They will help find strategies to enhance swisscovery’s appeal in research data options.

For the project, the abstract SLSP personas will be extended with practical user stories: A student/researcher wants to find and access resources/datasets. As Connectome also offers information on people and projects, the need to network with other researchers is likewise addressed. The interest in ‘grey’ data and open educational resources (OER) is treated as special cases.

To recruit participants covering several domains, a call was published on several platforms. The user workshops would use mixed methods such as focus groups, design thinking, and usability walk-throughs, as well as an open approach.

The idea was to have a series of workshops. A hypothesis (1) would be created in the first one, and an evaluation would be presented in the last one (2) as shown in Figure 3. The first workshop has already taken place, focusing on understanding and prototyping research data discovery needs in swisscovery. Several researchers with cross-disciplinary backgrounds (computer science, health science, and finance and law) participated in the workshop. Common tasks to find research data were detected, covering topics such as raw data versus secondary data, data impact, full-text access, data granularity, and reference management. Afterwards, empathy mapping was performed to highlight critical points in the discovery process. All the workshops will result in the creation of initial conceptual wireframes and designs in harmony with the acceptance criteria of a discovery prototype.

Future Challenges

The project itself faces several challenges. Regarding the data perspective, the ‘Swissness of data’ can create problems since researchers are usually more interested in an international focus. Furthermore, there is a difference between standardised data aggregation and standardised data curation. SLSP has an established community with which it can organise data curation within a framework (e.g. international cataloguing regulations). However, Switch harvests data and can merge it in a standardised way but has little influence on how the providers offer the data. Finally, there is an overlap between Connectome and swisscovery in the provision of textual data, which raises questions about entity matching and whether the duplication of work can be avoided. Furthermore, there is also a difference in the business model. Connectome is already a few years old but still has the character of a flexible project, whereas swisscovery is the established main service of SLSP and its community and is, therefore, more firmly anchored. Although both institutions operate in an academic environment, the specific wishes of the shareholders may differ, which can be reflected in different visions for the project.

Much depends on the specific design of a user interface for a web service. For swisscovery to be able to offer its users a multifaceted service (e.g. resource sharing), the interface is already heavily filled with information, and there is a risk of overloading with the further integration of resources such as Connectome. Conversely, the integration of bibliographic data in Connectome can lead to such data being lost alongside project information and datasets, resulting in too much data scattering. Usually, the scope is important for information retrieval; in other words, every data type may need its special discovery platform. Finally, external dependencies can also pose a risk to the project. Specifically, swisscovery is based on the Primo VE system from the provider Ex Libris. A next-generation discovery experience is currently being developed for this system which will be reflected in a new UI, and SLSP is dependent on this development.^[9] As the project runs for some time, it can lead to further unexpected changes, e.g. disruptions through technological leaps in artificial intelligence.

The Case of the Swiss National Library

Within the libraries of Switzerland, the Swiss National Library occupies a prominent position, as its legal mandate places high demands and has numerous requirements. Due to these special circumstances, the library has not yet been integrated into the SLSP network. The library’s collections include a vast array of materials, covering a wide range of subjects, which reflect the cultural, historical, and academic diversity of Switzerland. The first incorporating steps have been taken since the end of 2023, and the holdings of over 500 academic libraries from swisscovery and the Swiss National Library have been searchable in a new search portal, Recherche Patrimoniale. This advancement has been made possible with SLSP’s service SLSP Bibliographies which uses its framework to create special views that only provide bibliographic information.

SLSP Bibliographies: Meeting the Needs of Exposing Heritage Collections

Through a service called SLSP Bibliographies, swisscovery not only aggregates and displays cultural data but also offers specialised views. It is designed to showcase bibliographical references and specific collections without direct access to services; instead, it redirects users to swisscovery or other search platforms. The feature is particularly useful for creating themed collections or views focused on specific regions or topics. This service defines not only the more flexible configuration options but also certain limitations such as the duty to hide fulfilment-related UI elements, as these views should not compete with the regular local views of the swisscovery network which need to follow many pre-defined standard settings by SLSP. These standards are handled via SLSP’s Master Template, where all discovery configuration parameters and differentiated, mandatory, and optional settings are documented as an orientation for all swisscovery libraries. In addition, a common usage of display fields in Primo is defined in the Master Template. For the bibliographic views, some flexibility is granted on the view level configuration, such as free choice of labels used for the specific view, facets of choice, brief record and full record displays, and many other configurations, while the parameters on the institution level need to stay untouched. Figure 4 provides a snippet from the SLSP discovery services definition, where the regular local views and the SLSP Bibliographies’ views are handled differently in terms of allowed configurations.

Examples of SLSP Bibliographies include views created and managed by the Bibliothèque cantonale et universitaire de Fribourg for the cantonal bibliography and ETH Zurich for international collections, such as the Thomas Mann International search portal. The first example of the Bibliographie fribourgeoise view,^[10] showing the bibliography of the canton Fribourg, lists printed and audiovisual documents with a Fribourg theme, as well as literary, musical, and artistic publications by Fribourg authors. The research portal TMI Research of the Thomas Mann International network provides access to extensive archives from various institutions, including Buddenbrook House/Heinrich und Thomas Mann Centre in Lübeck, Monacensia im Hildebrandhaus in Munich, Thomas Mann Archives of the ETH in Zurich, and Thomas Mann House in Los Angeles.^[11] It offers resources such as manuscripts, letters, photographs, and personal libraries of Thomas Mann and his family, with many materials digitised and searchable online. Access includes specialised databases, photo archives, and a comprehensive collection of works related to Thomas Mann’s life and influence. The data from the various sources was harvested to ETH and separated from the regular inventory as most of these resources do not provide direct access to the sources, unlike to the rest of the resources found in swisscovery. This view served as a pilot for SLSP to establish the SLSP Bibliographies service through close exchange with ETH, but also with Ex Libris to accelerate certain developments which were crucial for the configuration. Another example is the Recherche Patrimoniale,^[12] a project between the Swiss National Library and SLSP (Figure 5).

Swiss National Library and Recherche Patrimoniale: A Collaborative Effort

The Swiss National Library and SLSP collaborated to create the Recherche Patrimoniale platform, which offers the public easy and extensive access to the participating libraries’ collections. SLSP manages the operation of Recherche Patrimoniale, ensuring that SLSP standards are applied to both data processing and presentation (Swiss National Library, 2024). Although the Swiss National Library is not integrated into the SLSP network, the collections from both institutions have been made accessible through the Recherche Patrimoniale portal, uniting resources across two different Alma environments by harvesting the National Library’s data via OAI-PMH into the Alma instance of the SLSP Network Zone. Like in ETH’s example above, this data has been excluded from the regular swisscovery search scopes to avoid mixing titles into the results where no direct access to the resources is possible for the swisscovery users. The titles are, therefore, only searchable in the dedicated view in the SLSP network zone. This specification means that the search portal redirects end-users from this view to the original search platforms swisscovery and Helveticat, as shown in Figure 6.

Future Challenges

The SLSP Bibliographies concept faces organisational challenges, including varying requirements and knowledge levels across libraries, highlighting the need for thorough requirements elicitation. Technically, there are limitations in excluding external data from regular searches, with only three data sources being allowed for exclusion and no post-login data exclusion in the network zone view swisscovery. The ability to configure only at the view level further restricts, for instance, settings related to multilingualism for a specific view or the creation of an additional search index. Moreover, operating within a complex and large network introduces performance issues when local search indexes are added. Another challenge is ensuring that digital objects which are stored in Alma can be shared across the network without inadvertently affecting other institutions’ local search scopes in their local Primo views. Another possibility could be to refrain from using digital storage in Alma but to use the solution of the configuration of remote repositories in Alma.

Looking ahead, there are opportunities for further collaboration, both within Switzerland and internationally. The Recherche Patrimoniale and the Thomas Mann International view could serve as models for future initiatives that bring together even more patrimonial libraries and cultural institutions within the SLSP framework. These future collaborations will require careful planning, resources, and a clear understanding of each institution’s unique needs and legal requirements. Also, the technical limitation regarding the separation of certain titles deriving from external data sources, as described above, needs to be assessed.

When considering fully integrating patrimonial libraries, such as the Swiss National Library, into the swisscovery network and not only bringing them together in a joint discovery platform in the future, several critical aspects must be evaluated. Patrimonial libraries, unlike public and academic libraries, are legally required to preserve and make available certain collections, which can include bibliographical metadata that forms part of the cultural patrimony. In the swisscovery network, catalogue data can be modified by all participating libraries, raising concerns about data sovereignty and the integrity of these collections. Additionally, the integration of APIs and interfaces with external data sources needs careful planning to ensure compatibility and security. User data management also presents challenges as swisscovery users are connected via Switch edu-ID and managed centrally, but cantonal libraries include a broader user base, such as pupils and elderly individuals, necessitating tailored solutions. Finally, legal aspects, including terms of use, must be clarified, as several cantonal libraries operate under different terms than those harmonised across the SLSP network. Addressing these issues will be crucial for a successful and sustainable collaboration.

Conclusion and Outlook

This contribution aims to provide an overview of the data available in swisscovery and its handling by SLSP and to deliver insights into the current state of the Connectome project and the National Library case studies. Although both projects have different focuses, they have a major commonality, namely discovery. One challenge is to ensure that the user is not overwhelmed by the information being offered while at the same time maximising the use of resources. As Bertino (2024) points out, intelligent suggestion systems can offer help here, even if they harbour risks (e.g. confirmation bias and popularity bias). A current example is the Research Assistant from Ex Libris. This AI-powered tool enhances users’ search experience in swisscovery by allowing natural language queries and providing summarised answers based on a comprehensive index of over five billion scholarly records from various publishers, aggregators, and repositories. The tool uses large language models to generate reliable responses, with sources displayed for easy verification and further exploration.

To improve the user-friendliness of swisscovery, SLSP commissioned the Swiss Institute for Information Science (SII) at the University of Applied Sciences of the Grisons to conduct an independent usability study in 2013.^[13] The findings of this research will be continually utilised to enhance swisscovery. They can also be used for neighbouring initiatives that have discovery components (i.e. patrimonial view and Connectome).

Ongoing collaboration between major Swiss players is crucial to keep pace with the evolving data infrastructure, and networking with loose networks such as OpenGLAM can also be beneficial.^[14] Additionally, generating impact and reaching target groups for digital objects requires innovative approaches, such as leveraging visual platforms such as Instagram or TikTok, as Marika Sarvilahti’s contribution demonstrates (2025).

In summary, swisscovery’s efforts to preserve and make accessible Switzerland’s cultural heritage are crucial for supporting research and education. Through services such as SLSP Bibliographies and projects including Connectome and Recherche Patrimoniale, SLSP, the organisation behind swisscovery, continues to enhance access to the country’s rich cultural data, benefiting researchers, scholars, and the public. In the future, ongoing collaborations, thoughtful customisations, and strategic outreach will be key to addressing upcoming challenges and ensuring the continued success of swisscovery.

References

Abstract

The Swiss Library Service Platform (SLSP) was established in 2017 to consolidate and centralise library data and services across Switzerland. This article examines two projects that illustrate SLSP’s role in managing and disseminating cultural heritage data: the Connectome project and the collaboration with the Swiss National Library. The Connectome project, in partnership with Switch, aims to promote open data practices by linking diverse metadata into a knowledge graph, enhancing discoverability through artificial intelligence-driven recommendations. The collaboration with the Swiss National Library focuses on integrating the library’s vast collections into the swisscovery platform, creating a unified search experience. Both projects underscore the challenges and opportunities in managing cultural data, particularly in terms of data integration, user experience, and discovery. The article especially highlights the importance of usability in these initiatives, as SLSP strives to ensure that users can effectively navigate and access Switzerland’s rich cultural heritage without being overwhelmed. Thoughtful customisation, ongoing collaboration, and enhancements are crucial for overcoming future challenges and maximising the impact of these discovery platforms.

Keywords

Cultural heritage; Usability; Discovery; Metadata integration; SLSP