The Muninn project is an academic research project creating Linked Open Data from First World War Archival, Library and Museums sources. The aim is to take digitized documents, extract the data using computing power and turn that information into structured Linked Open Data that is published under a creative commons license and a part of the Linked Open Data Cloud. Inter-Linked to other well known Linked Open Data Cloud datasets such as Linked Geo Data, Geonames and Dbpedia, Muninn data can then support further historical and semantic web research by providing real-world information in a format that is easy to access through URLs, data dumps and a SPARQL server.
Some of the current questions that the Muninn team has been working on is how to consume historical Linked Open Data with tools other than graphs and spreadsheets. One solution that we are presenting here are as objects in virtual worlds.
Books surround us and are everyday cultural heritage objects. With the availability of large amounts of digitized books with linked open data cataloging data, the instantiation of their contents within a virtual world is straightforward.
The content of the books themselves can be displayed within the virtual world. Here, The Count of Monte Cristo, a 1834 novel by Alexandre Dumas, is randomly left on a house table for visitors to peruse.
Since Linked Open Data gives us the ability to locate books based on year, language, culture or context, their automated insertion into historical scenes is possible without creating an historical anachronisms. Hence, this copy of The Three Musketeers would not appear in a simulation dated earlier than 1844, the year of the first publication of the work.
This allows the creation of opportunities for the serendipitous discovery of new relevant books to the context of the simulation without incurring the programming overhead normally required to do.
Social Networks data about the owners of objects can also be used to further enrich the world in ways that are contextually appropriate.
Only a limited subset of the creative works curated by a Museums are on display at any given time. Works by lesser known artists are not always given the attention they deserve nor do exhibitions cater to all tastes. Through the use of Linked Open Data API that are provided by some Museums such as the Rijksmuseum creating specialized, thematic or genre specific virtual exhibits is straightforward. It also allows for the serendipitous use of creative works in virtual environments for an increased realism while communicating the esthetic of a certain historical period.
We can push this approach further by using the content of other Museums, such as the Cooper–Hewitt Smithsonian Design Museum, to programmatically theme the generic elements of virtual locations with the styles appropriate to an era.
The Muninn project has created linked open data sets of Central and Allied trenches over the western front extracted from original trench maps of the period using image processing algorithms. The availability of online linked open data maps catalogues enables the automated generation of this data and its visualizations in 3D virtual worlds. This allows us to experience locations that no longer exist and enrich scholarly study of past events.
Ontologies such as the Muninn Graves ontology record the location of graves in different cimeteries. Their coupling to existing data sets and visualizaiton engines can show the evolution of a cemetery and the location of unmarked graves. The Fort Massey Cemetery has been in Halifax since the 1750. It contains the graves of soldiers from different eras, not all of which have a grave makers. Temporally enabled linked open data makes it possible to visualised the same data set at different times to plan an archelogical dig, research an ancestors’s grave or navigate the modern day cemetary.
As an open format, Linked Open Data can be used for different applications. Here, triples about the Battle of Vimy Ridge that occurred during the great war are used to create a solid model of the battlefield that can be 3D printed. This specific piece is a representation of the eastern slope of Vimy Ridge in 1917 with German communication trenches clearly visible. While the printing raised-relief maps in not new, the use of linked open data to print features such as trenches that no longer exists is a novel way of experiencing lost heritage sites.
The inherent power of Linked Open Data, and BIG Linked Open Data, is the ability to derive value from data that would otherwise be discarded as useless. This plot shows the details of Royal Navy ships tracks during WW1 as derived from the Zoom Universe old weather data set. While the Old Weather focuses on transcription projects and Muninn on the First World War, neither are experts on weather data and the quality of the weather data remains undocumented.
The W3 Sensor Ontology allows us to link the ship’s tracks with their weather observations as well as the original sensing instrument. In this case an instance of the original barometer / thermometer combination instrument is available at the National Maritime museum in Greenwich. Not on display, the existence of the artifact would not normally be known if not for the museum API.
Linked Open Data makes the data not only available but documents it in a way that is useful to climate scientists, who care about the process used to record the weather data, naval historians who care about the documented proof of the position of the ships and to cultural heritage curators who about the underlying documents and artifacts from which all of the information flows.
In the coming year Muninn will continue generating GIS data about the Great War and working on novel data consumption approaches. Two applications of interest are the use of Great War maps and aerial pictures to create unexploded ordnance risk maps and data extracted from the personal letters of Great War soldiers to extract social network information in a manner similar to the Linked Jazz Project.