Introduction to Sampo-UI

Joren Six · GhentCDH

Core facility

Service provider

Digital support for
humanities researchers

My focus: IT services

Research centre

Research & collaboration

Digital methods for
humanities research

Education

Teaching & training

Digital skills for
students and researchers

GhentCDH projects

  • Roman coin finds in a research database
  • Linguistic annotation of scanned Greek papyri
  • Belgian artists’ social networks: Navez
  • Ghent heritage archives explored through maps
Bestor project interface
Bestor
Cinema Belgica project interface
Cinema Belgica
Gent Gemapt project interface
Gent Gemapt
NuMAD project interface
NuMAD
Bestor faceted search interface
Cinema Belgica faceted search interface
Diplomata Belgica faceted search interface
EVWRIT faceted search interface
Gent Gemapt faceted search interface
Novel Echoes faceted search interface

Repeating patterns

Different research questions. Repeated work in how data is presented, stored and queried.

  • Custom faceted interfaces and search logic
  • Custom data visualisations
  • Custom data query interfaces
  • Custom data export functionality

Data islands, without a common strategy.

Repeated development creates overhead A clockwise cycle of building, adapting, and maintaining custom tools starts again for the next project. The recurring work creates development and maintenance overhead. Repeated work Development & maintenance overhead BuildAdaptMaintain Next project Same needs. Another implementation.

Why Sampo-UI?

A shared foundation for recurring needs.

  • Configurable faceted interfaces
  • Configurable data visualisations with plugins
  • A standards-based query interface: SPARQL
  • Shared data export functionality

Reuse the framework. Configure the project.

A shared framework supports reuse across projects A clockwise cycle of configuring, adapting and extending, shared maintenance, and moving to the next project. The shared framework carries improvements forward. Shared foundation Reuse & shared improvements ConfigureAdapt andextendSharedmaintenance Next project Next project. Same shared foundation.

How Sampo-UI works

The Sampo model separates the data service from the user interface.

Sampo model: separate data service and user interface The Sampo-UI web frontend sends SPARQL queries to an independent linked data service and receives results. Researchers can also query the same service directly for further analysis. Linked data service SPARQL endpointShared data and ontologies Sampo-UI web frontend Filter → analyseMaps · tables · networks SPARQL queries Results Direct data access → further analysis One data service.Multiple ways to use it.

Semantic Computing Research Group · Aalto University / University of Helsinki
Reference: Digital humanities on the Semantic Web: Sampo model and portal series

AcademySampo: People main perspective with search facets and table results
FindSampo: Finds main perspective with search facets and table results
Mapping Manuscript Migrations: Manuscripts main perspective with search facets and table results
OperaSampo: Performances main perspective with search facets and table results
BookSampo: Novels main perspective with search facets and table results
LetterSampo Finland: Letters main perspective with search facets and table results

Sampo-UI and GhentCDH

From a local fork to contributions in the shared framework. Both CLARIAH funded.

Sampo-UI development and GhentCDH contributions Sampo-UI development at SeCo continues along the main branch. GhentCDH forks the project, modernises its architecture and software stack, and improves modularity. The fork has merged back into upstream development. SeCo · Helsinki / AaltoSampo-UI development Fork Merged upstreamShared development GhentCDH modernisation Architecture · software stack · modularity Contributions return to the shared framework

NuMAD: public SPARQL endpoint

The data behind the portal
is directly queryable.

  • Ontop exposes the relational data through SPARQL.
  • The portal and external analyses use the same endpoint.
  • The mapping determines which data is published.

Open the endpoint interface

Shared vocabulary: Nomisma

Use an established ontology
to describe coin data.

  • Shared concepts for numismatic research
  • Consistent meaning across collections
  • A foundation for linking and reusing data

Explore the Nomisma ontology

NuMAD live demo

NuMAD portal showing coin filters and a map connecting mints to coin finds.

Open NuMAD · Filter coins, switch views, explore the results.

Data editing needs its own environment

Sampo-UI provides the public search and analysis interface. Data entry and editing happen in a separate application.

  • Create and correct records in a dedicated data management environment.
  • Manage validation, permissions and editorial workflows there.
  • Publish the data through a SPARQL endpoint for Sampo-UI to explore.

Plan for both: a data editing environment and a Sampo-UI portal.

Maintaining and editing a graph database

  • Graph databases require specialist expertise.
  • The supporting software ecosystem is less rich.
  • Expertise for long-term support is limited.

Federated queries: multiple endpoints

  • Querying multiple distributed endpoints introduces external dependencies.
  • All is well until it is not.

Link to external data, but cache locally for reliability and speed. Avoid dependencies.

NuMAD architecture

Within Ghent University, the NuMAD Roman coin finds database connects via SQL to Ontop and via a two-way SQL link to the Data Management Web Interface. Ontop serves the Sampo-UI Web View via SPARQL. The two interfaces serve external users and researchers via HTTP, respectively.

An existing relational database, exposed through Ontop and explored with Sampo-UI.

Conclusion

Sampo-UI provides a strong data-visualization toolbox.
and nudges projects to publish linked data.

Thank you

Questions?
joren.six@ugent.be