The ensembles of the Army museum
In recent years, the Legermuseum (Army Museum) has been actively involved in registering and digitising its collection. A large part of this collection is accessible through the Collectie Informatie Centrum (Information Collection Centre) or via the museum’s website. However, just registering and digitising objects is not enough. Objects will only come to life if they are placed in context, when visible links are established and when the public is invited to browse through the collection and contribute their own stories. This is something that will require a substantial amount of work behind the scenes.
Between 1999 and 2005, the Legermuseum was able to solve its registration backlog thanks to the Deltaplan Inhalen Achterstanden Registratie (backlog recovery delta plan), which was backed by the Dutch Ministry of Defence. Approximately 400,000 objects, including books, engravings, drawings, photos, uniforms and weapons were cleaned, photographed and repackaged. Thanks to this basic registration process, the museum presently knows exactly which objects are in its collection. At the same time, the Legermuseum participated in the ‘Geheugen van Nederland’ (Dutch Collective Memory) project launched by the Dutch Ministry of Education, Culture and Science and comprised the creation of a database of digital images relating to the history of the Netherlands.
Annet Ruseler, the Legermuseum’s interim Knowledge Transfer Manager, was involved as project leader in both projects. ‘The ‘Het Geheugen van Nederland’ project marked our first venture into the world of digitisation. The Legermuseum had to start off somewhere, but where? This project allowed us to show the diversity of our collection using 7,000 objects, while simultaneously focusing on the preservation of the collection. We started looking at the sub-collections and evaluated which were the most vulnerable. For example, we set out by digitising our collection of glass negatives, which included various unique objects. These images can now be viewed at the Collectie Informatie Centrum (CIC) as well as on our website. We have stored the originals in a safe location, with the result that it is no longer necessary to visit the study room to view them. After the project was completed, we continued digitising our objects. At present, the Legermuseum has made all its engravings and drawings available in digital form. However, this is not an option where our photographic material is concerned. We own approximately 75,000 photographs, not counting various valuable books that contain engravings, printings and personal documents. Consequently, we have to set priorities. At the moment, we are mainly guided by the demands of our internal and external customers, publications, exhibitions and our website. Should someone request pictorial material not yet available in digital form, we will digitise it. Should this concern an engraving from a very valuable book containing various other striking pictures, we will have them digitised at the same time. In this way, the book will not have to leave our premises for subsequent digitisation purposes. All of this means our digital records are gradually expanding.’
Stories
‘At present, the foundation is formed: we know exactly which items we own and actually have picture of many objects. The Legermuseum intends to tell stories. It doesn’t want to limit itself to presenting individual objects. Its objective is to present ensembles, in which objects gain significance through their context and their interrelation. The Legermuseum is able to collect all sorts of material from its cabinets and to tell a story about a particular person, using everything including his uniform, his notebook, his weapons and the photos he took or is depicted in and the books in which he is referred to. Although it may be easy for the custodian or (amateur) historians to show the link between objects and to place objects in a broader context, this task is beyond the expertise of the average visitor. The Legermuseum strives to tell visitors the complete story by presenting objects as part of an exhibition and to encourage visitors to browse through all the information. We want to kindle new ideas in people: you currently find yourself at such and such a point in our collection, but we have significantly more material on this topic. Have a go and log on – you have a lot more to see. It’s like surfing from one web page to the other. Virtual visitors must be offered an efficient search structure.’ This is where the RNA project comes into play.
The RNA project
The Legermuseum features a collection database called Adlib. This enables you to do a search on specific military terms that are in many cases not hierarchically interlinked. The RNA project helps the Legermuseum to ‘automatically’ create these structures and to superimpose an umbrella of less specific terms on them. In this way, the objects of the Legermuseum are placed in a (broader) context and can be presented as a coherent whole. In a bid to provide the search system of the Legermuseum with a more coherent structure, it has resorted to using WordNet, a well structured list of keywords. This is an American word-based system, a structured ‘dictionary’ (not just related to military topics) in which all terms have been assigned a well-defined place within the word hierarchy.Although the terms in WordNet have been translated into Dutch, their descriptions are in English. As a result, project WordNet has been linked to two digitised Dutch dictionaries for the purpose of the RNA:
- Verschueren, a Belgian dictionary that has been digitised into XML format. This dictionary was made available by the Bibliotheken (Dutch libraries), another partner in the RNA project.
- Landolt, a military dictionary dating from 1860. Part of the Legermuseum’s collection, this dictionary was also digitised in collaboration with the DBNL.
The Dutch descriptions from these dictionaries have been linked to the terms in WordNet. This allows a reference structure to be created that serves as an umbrella for the key word systems used for the Legermuseum’s database. By linking individual terms and clusters of keywords in the database to the reference structure, a coherent whole is created. This offers three key benefits:
1. From their location in the umbrella structure, end users can penetrate various locations in the database, return to the umbrella structure and descend from another umbrella rib, without actually being aware of this deviation (see ‘Browsing the database’). In this way, users of the online catalogue are offered a context as well as suggestions for new ideas
2. By creating semantic links between terms (i.e. looking into the meaning of those terms, see ‘The right link’), the Legermuseum is now able to make more efficient automated (machine-readable) use of the reference network. In addition, it can do so by adding new articles to the database and making them searchable, or to create links with other knowledge systems that are also based on the ‘semantic web method’. This may well be the greatest value added by the project.Browsing the database
If you are searching, for example, for a Dutch soldier, person X, who served under Napoleon and William I, and wish to know how it was possible for a Dutchman to fight under a French flag, you can click on the French army during the period 1812-1814 or on the Kingdom of the Netherlands. You may end up at the Russian campaign or the uniforms that were worn and the weaponry that was used in those days.
3. If the Legermuseum were to create a reference structure using a manual method, it would take years to complete this process. In other words, the RNA project saves a lot of time. Annet Ruseler: ‘There are many Dutch museums of various sizes focusing on military-historical heritage. Like the Legermuseum, most of these museums use Adlib, which will make it possible in the future to access these databases through one and the same reference network.’
The right link
When data in various databases is linked on the basis of an automated process, the risk exists that the computer might accidentally link terms with the wrong significance. For instance, the Slag bij Waterloo (Battle of Waterloo) from 1815 may be linked to the term ‘slag’, which is interpreted as a 'movement' rather than a ‘campaign’. However, this can be prevented if the computer system is programmed in such a way that it takes into account terms higher up and lower down in the hierarchy during the linking process. For example, if the underlying term ‘slag’ is ‘hockeyslag’ (hockey stroke), the term ‘slag’ will be interpreted as referring to a ‘beweging’ (movement), and consequently the system will know that no link must be created to the Russian campaign. However, if the underlying term is ‘campaign’, the significance will be understood as ‘battle’, in order for linking to take place.
Creating combinations
The RNA project offers a tool with which the Legermuseum can solve its entire information-disclosure puzzle. Annet Ruseler: ‘We try to combine as many tasks as possible. I already mentioned the example of digitisation: whenever an image is requested that is not available in digital form, we will have it digitised and include other images from the same source where possible. If we have photos taken for an exhibition, these photos will be registered and stored digitally so that we can use them on future occasions.If a curator is preparing for an exhibition or publication, we will ask him or her to evaluate the information in the database: have the proper terms been used, are there any instances of duplication, and is there any knowledge to add? The Legermuseum presents a large variety of antique hand-guns in its ‘Bloedmooi’ (To die for) exhibition. These firearms form part of our collection and have been registered in Adlib. Using a predefined search query, the relevant data can be retrieved from the database and added to the website, accompanied by photographs and text descriptions. At the same time, the search query will be used to present the information during the exhibition, complete with the exhibit number.’
Involving the audience
‘The information made available through the website is still largely determined by the Legermuseum. However, we want to offer visitors the possibility to browse through the information, to discover new things. We also want to offer possibilities for interaction, e.g. the ability to add new or to enrich existing information. We receive a lot of feedback with regard to topics such as the Netherlands East Indies and New Guinea. People come up with their own stories and photographs, and in some cases they have even created a website relating their personal experiences. It would be great to offer a platform through which people can add contributions. Who knows – maybe we could create a small, continuously expanding online community. There are many other ways to involve people with our collection. At the CIC – as well as virtually, on our website – we will set up a showcase with objects that we would like to know more about. We want to invite people to add information through a Q&A system.’ A new RNA project is in the pipeline that involves experiments such as social tagging. Users will be able to link labels (tags) to objects. In order to be able to process these labels in the database in a flexible way, identical terms must be used. If a contributor labels a photograph of the devastated inner city of Nijmegen with the term ‘bombardment’, and another contributor uses the term ‘bombardments’, they will both mean the same thing. However, two keywords will be created in the system that will not be automatically recognised as synonyms. In the short term, the RNA project will start experimenting with a combination of social tagging (the free association of keywords) and a controlled keyword list. Users will be able to enter a term, while the thesaurus administrator, for example, will decide where the term should be entered in the structure.
Thursday, August 16, 2007