Representing, Using, and Maintaining Military Historical Linked Data on the Semantic Web

The Second World War is the largest global tragedy in human history. It is extensively documented in historical sources, but this information is scattered in various organizations and countries, written in multiple languages, and represented in heterogeneous formats. Semantic Web technologies provide solutions for combining heterogeneous distributed historical information. By combining information from distributed sources it is possible to get a deeper understanding about the history than by studying the sources individually. This thesis explores the use of Semantic Web technologies for representing and modeling heterogeneous military historical information as Linked Data, with a focus on depicting the history of Finland in the Second World War. Harmonization and integration of military historical data from distributed sources are studied, while also investigating how to search, browse, analyze, and visualize the resulting Linked Data on web-based user interfaces. Maintenance of the highly interlinked set of graphs exposes new challenges and a solution to tackle them is presented. These topics are studied in the context of building the WarSampo information system. Linked Data and the event-based CIDOC Conceptual Reference Model are used together in WarSampo to achieve the interoperability of heterogeneous military historical datasets. Events are used as the glue which combines together information from various source datasets. The event-based modeling enables depicting the national military history narrative as data, which can be further enriched with the events of individual military units and soldiers. This idea is demonstrated in the WarSampo semantic portal, which consists of nine different perspectives on the data integrated from distributed sources. Each perspective provides a customized user interface for a certain part of the WarSampo knowledge graph, like war events, persons, wartime photographs, and places. The knowledge graph is published as open data and is a part of the global Linked Open Data Cloud. The WarSampo portal at http://sotasampo.fi is a popular service for citizens to study the wars, and to find out what happened to their relatives. It has been used by more than 660 000 end users, equivalent to more than 10% of the population of Finland. The proposed methods and data models are useful beyond the geographical and temporal scopes of this research. The aspiration behind the WarSampo project is that by making military historical data more accessible, our understanding about the reality of the war will improve, which also promotes peace in the future.