On the 28th of November 2024, I [1] attended the “Tutorial: How to create a Linked Open Data service and semantic portal for your Cultural Heritage data” organized by the Semantic Computing Research Group (SeCo) and live-streamed from Eespo, Aalto University. I was interested in participating in this initiative as I am a PhD student in Cultural Heritage in the Digital Ecosystem at the University of Bologna, in joint supervision with the University of Helsinki (Computational History Group), and my research has a specific focus on RDF data and Semantic Web technologies. I got to know about the Sampo national semantic web infrastructure and its “Sampo series” of Linked Open Data (LOD) services and portals a couple of years ago while conducting individual research, and it was very easy to realize the long-term effort behind its development, which led to a network of well-integrated services for structuring and publishing RDF data. In particular, I decided to follow this tutorial to gain a deeper technical understanding of the whole framework, with the final aim of understanding how to develop a Sampo portal and evaluate the viability of exploiting this network of services for making my research data available online.
A PIPELINE FROM RDF MODELING TO THE CREATION OF A SAMPO PORTAL
The series of speeches started with a brief presentation of the SeCo research group and its activity concerning the Sampo development and long-term maintenance, followed by a short but detailed introduction to the concept of the Semantic Web, with particular attention to explaining its principles and potentialities. The following presentations were organized so to provide an overview of the full pipeline for transforming structured data such as CSV or JSON in LOD to be presented in a Sampo portal, from the very basics to such topics as data enrichment, the creation of a SPARQL endpoint, semantic data validation and NLP applications for improving informative content of RDF graphs.

The section specifically devoted to the creation of a Sampo portal started at the beginning of the last quarter of the full tutorial, which lasted around two effective hours and concluded with an overall presentation of some of the Sampo portals developed on various topics, and quick live demos of searching, browsing and visualization functionalities. It was then highlighted how the variety of the informative content shown and the navigation possibilities could serve a broad and diversified target audience including regular users, researchers, journalists, and data managers.

The overall feedback that I can give about the project itself is enthusiastic: the commitment and effort devoted to it is evident and the results are impressive, since the SeCo group practically got to develop a framework that makes it concretely possible to exploit the Semantic Web opportunities fully. The tutorial successfully provided a complete overview of a pipeline for producing, enhancing, and sharing information as Linked Open Data, which I believe was particularly useful for expert users of digital technologies in general, who are new to the Semantic Web in particular. Thus, I would identify this latter category as the real target of the tutorial. Indeed, I believe that the structure of the presentation could have been perceived as too technical by humanities data experts new to digital technologies, and as too introductive for the LOD community, who might have expected some hands-on activities and didn’t need wide introduction about the enhanced informative potential of converting data to be published for the Semantic Web.
As a pitfall, I believe that through the sequence of presentations preceding the effective tutorial to build a Sampo portal wasn’t always intuitive where a global overview of the technical possibilities for accomplishing a task of the pipeline ended and where (and whether) a suggested well-defined workflow started. Thus, the degree of freedom in the data modeling, conversion, cleaning, enrichment, and validation phases might have made it practically complex for a non-LOD-expert user to select tools among the reviewed alternatives to practically prepare the data for a Sampo project. However, this aspect probably didn’t affect Semantic Web-acquainted users.
Personally, even as someone who regularly deals with Linked Data, I appreciated the first part of the tutorial as a useful state-of-the-art review. The following part helped me better understand Sampo tools’ structure and how they can be exploited jointly. I also gained the impression that creating a portal should be relatively easy and worth the effort.
All the resources concerning the event are generously shared, including the video recording and the presented slides, which also contain links to step-by-step instructions for creating a new portal. This is also in view of the ongoing project that SeCo is conducting to connect all the Sampos in a wide and consistent network of interlinked resources.

CONCLUSIONS
In conclusion, I believe that the Sampo framework represents a complete and well-integrated resource for maximizing the exploitation of Semantic Web technologies on a variety of research data. The tutorial was interesting, informative and complete, even if the introductory part was probably too long for the LOD-expert users. For this reason, although workflows for realizing a portal are already made available online and the links to those were also shared in the slides, I think that dividing the tutorial into two appointments could have provided the opportunity for splitting the audience according to their expertise level, so to provide a maximally useful experience to both of the groups. In particular, a second online interactive tutorial could have been scheduled, devoted to hands-on activities for incentivizing expert users to share their data through a Sampo portal, increasing the chances to broaden the overall network. However, I found the presentation very useful, since it provided me with an interesting refresh and some new ideas concerning pipeline tools, and convinced me to individually keep exploring Sampo’s potential for my research interests.
[1] Text: Arianna Moretti. Moretti is currently pursuing her PhD in “Cultural Heritage in the Digital Ecosystem” at the Universities of Bologna and Helsinki.
Images:
Title image: Visualization of LOD-cloud and The Forging of the Sampo (Akseli Gallen-Kallela, 1893). Image provided by Aalto SeCo Group.
[Fig.1] https://seco.cs.aalto.fi/events/2024/2024-11-28-sampo-tutorial/ahola-rantala-2.pdf
[Fig.2] https://seco.cs.aalto.fi/events/2024/2024-11-28-sampo-tutorial/rantala.pdf
[Fig.3] https://seco.cs.aalto.fi/events/2024/2024-11-28-sampo-tutorial/leskinen-leal.pdf