The WarSampo.fi portal opened at the end of 2015 and attracted twenty thousand visitors over just a few days.
“The amount of visitors evened out after this initial surge, but WarSampo is still used a lot,” says the project's leader, Professor Eero Hyvönen from Aalto University's Semantic Computing Research Group.
An enormous amount of material has been gathered to be published openly to all members of the public under WarSampo. An extensive group of cooperation partners has participated in the provision of information and the creation of this service. The Finnish Wartime Photograph Archive, for example, provided the portal with data on 160,000 wartime photos and videos, while the National Archives of Finland contributed data on, among other things, 95,000 front-line casualties and 26,000 war diaries. Some 3,300 articles from Kansa taisteli (The Nation Fought) magazine from 1957-1986, originally published by the Association for Military History in Finland and Bonnier, have also been linked to the portal.
Location information provided by the National Land Survey of Finland was used to generate a new historical location register (so-called ontology), which contains data on 35,000 places in the parts of Karelia Finland surrendered to the Soviet Union after the war. In addition, a register of 11,000 wartime events was created using various information sources.
In all, the portal contains over a million data items, such as Vyborg and Mannerheim. These are linked to each other to form a set of semantic web perspectives like places and persons, and together form an open data cloud. More than seven million connections have been created within the data. WarSampo's core data is also linked with external information sources like Wikipedia and the National Biography of Finland.
There is so much information that it would be difficult for a single individual or research group to manage. WarSampo makes it easier to process the huge volume of data, draw conclusions and produce analyses.
“The aim is to contribute to the construction of the Finnish identity by adding to our understanding of the most recent wars, and also to promote peace at the same time. I believe that the more we know and understand about war, the less there will be of it,” Hyvönen says.
Hyvönen has a personal interest in this subject. His father served on the front and his mother's side of the family comes from Värtsilä, which lies in surrendered Karelia.
Intelligent information coupling
The service provides WWII-related information to researchers, amateur historians, application developers and other online services. Users can delve into topics and combine nuggets of information into larger totalities. The material can be searched from the perspective of wartime events, places, army units, photos, Kansa taisteli articles, persons or casualties.
Part of the personal data, for example, comes directly from the WarSampo databases. In addition, the portal's AI will deduce what other information is available online on the individual. These other information sources help enrich the person's life story.
WarSampo utilises the semantic web data infrastructure as well as the Linked Data Finland publication platform and the associated tools created in 2003–13 as part of the FinnONTO project series, which was headed by Aalto University and the University of Helsinki.
As the information is being published openly, anyone is free to build their own application utilising the publication platform. Kinocompany Oy, for example, has developed the online service Sotapolku.fi, which collects wartime memories, photographs and information from citizens. Its core data on casualties was mined from WarSampo.
“WarSampo publishes all of its content as linked open data, and as a service than anyone can exploit also for commercial purposes,” Hyvönen notes.
Road signs at a fork in the road on the western bank of Kotikoski. Photo: SA-kuva.
Universities engage in cooperation
Eero Hyvönen works as a Professor at Aalto, in addition to which he serves as the director of the University of Helsinki Centre for Digital Humanities HELDIG.
At present, Hyvönen's research group consists of 12 people, the majority of whom work at Aalto. The University of Helsinki's HELDIG centre is, however, making a strong contribution to deeper research in this field and is in the process of filling eight new professorships associated with digital humanities.
“The two universities have been cooperating on this for a long time. Both have provided researchers for different semantic web projects, which has been very fruitful.”
The expertise of Hyvönen and his colleagues is in demand also beyond Finnish borders. A national project resembling WarSampo is being launched in the Netherlands. The country is also developing the giant Golden Agents project, which will link resources on the 17th-century Dutch Golden Age. Hyvönen is participating in the Golden Agents project's steering group.
Hyvönen is also involved with the EU-funded Reassembling the Republic of Letters project in which more than 30 European countries are participating. The project gathers and analyses correspondence from leading scientists from the 16th to the 19th century. The letters are studied to examine the spread of scientific thought in Europe. Voltaire, for example, sent almost 20,000 letters during his life. Together with Oxford University, Aalto University is developing linked data models, a portal solution and tools associated with correspondence and personal networks.
“We are contributing to the project as experts in linked data and are building a service, which will utilise the same Linked Data Finland publication platform as WarSampo. We are studying data visualisation associated with this topic in cooperation with Stanford University,” Hyvönen says.
A pioneer of this field
Eero Hyvönen started his career as an artificial intelligence researcher. Hyvönen says the field hooked him when he took part in organising the first Finnish AI research convention in 1984. This event evolved into our country's longest-standing series of conferences on data processing.
The AI researcher's career headed in a new direction in 2001, when the international W3C organisation launched the Semantic Web Activity programme, for which Hyvönen arranged a Finnish kick-off event. Since then, he has focused on researching and developing semantic web infrastructure and applications.
Hyvönen has published almost 400 articles and books, and he has been recognised with several international and national prizes, such as the State Award for Public Information.
“Intelligent web technology is a really interesting subject. Nothing human is alien to us. The semantic web can be utilised to benefit research in almost any field of science and life. Our inspired research group and dozens of data-producing cooperation partners are due a big thank you,” a delighted Hyvönen says.
WarSampo is one of the projects Aalto University is implementing to mark the centenary of Finnish independence. It has been nominated for the international EU Prize of Cultural Heritage.
Bomb aimer of a Soviet-made Tupolev SB-2 bomber, Captain Veikko Härmälä on submarine
detection patrol over the Gulf of Finland in summer 1944. Photo: SA-kuva.
Semantic network materials from culture to law
“The semantic web presents matters in a commonly agreed format, which is understandable to computers and enables machine logic to find items, and deduce possible links between different items,” Professor Eero Hyvönen says.
When all information producers utilise the same, jointly agreed data models and concepts, it is possible to combine data without mixing it up.
The infrastructure of the “Internet of Meaning” can be applied in the processing, management and publication of a multitude of different data contents. The WarSampo service, for example, is a continuation of the Kulttuurisampo.fi, Kirjasampo.fi and Matkailusampo.fi pilots, which were developed at Aalto University.
“Nowadays, the Kirjasampo.fi service is maintained by public libraries and it recorded 1.6 million visits last year. The system hosts the entirety of Finnish literature as a semantic web, containing linking of almost unrivalled richness anywhere in the world,” Hyvönen says.
For its part, the semantic Finlex (data.finlex.fi) compiles as linked open data the central part of Finnish legislation as well as case law from the Courts of Appeal and the Supreme Court of Finland that has been published by the Ministry of Justice. Intelligent legal services are being developed on the basis of this service. For the first time, semantic Finlex provides researchers access to Finlex material as machine-readable data as well as tools for analysing legislation and court cases through computational means.
This article is originally published in Finnish in the Aalto University Magazine issue 19 (issuu.com), April 2017.