Professor Peter Eklund talks about the development of the Virtual Museum of the Pacific – one of the first projects run out of the recently established the Centre for Digital Ecosystems, at the University of Wollongong. He and Peter Goodall built the website using Web 2.0 principles so that people of the Pacific could access their own cultural heritage collected by the Australian Museum since its inception in 1845. In this article, Professor Eklund gives CAN readers an insight into the technical issues that had to be considered and some of the challenges the virtual museum will face in the future.
Late 2006, I pitched the idea of a Virtual Museum of the Pacific to the Australian Museum. I had been working for many years developing a navigation paradigm using concept lattices. My PhD student Jon Ducrou and I had developed and evaluated several prototypes but what we needed was a real collection to work on. Australian Museum Director Frank Howarth and then Deputy Director Les Christidis were keen to do something “experimental” in the Museum and Web space, over and above their existing corporate Web-site efforts that are themselves innovative.
We applied for a University of Wollongong new partnership grant which paid for a DHTML prototype and allowed us to encounter first hand some of the issues that would later become central to the successful development of the Virtual Museum of the Pacific. On the back of this, in the following year 2008, we finally received backing from the Australian Research Council (ARC) for the Virtual Museum project.
I decided that we needed a very experienced hand to manage the project so I hired Peter Goodall who had nearly 30 years experience in the IT industry, Peter had worked in start-ups in Silicon Valley, for IBM in the US and most recently at Objective Corporation. Peter’s job was to work in the Museum on the project using an “insourcing” model of customer engagement. It allowed the Virtual Museum to be owned by the Australian Museum – not driven by outsiders.
An early issue for the Australian Museum (AM) and the research team was what would be in the Virtual Museum? The Pacific collection seemed a good starting point for a number of reasons but the collection was vast, containing nearly 60,000 objects of which only a small subset could be sampled.
Peter and I decided on an early increment that would get the content ball rolling. We proposed to sample 10% of the objects (about 40 objects) before Christmas 2008, we argued that this would be used to measure the overall time cost to the museum for processing 400 pacific objects. It also started the object selection process. The first 40 objects were from diverse locations, made of many different materials and represented a broad range of domestic and cultural objects. This was ideal for navigation using concept lattices. Melanie Van Olffen, our collaborating anthropologist at the Australian Museum, established this selection pattern and it continued to the project conclusion with great success.
Another piece of detective work for the research team involved understanding the corporate taxonomy used by the Australian Museum to classify their collections. The metadata used by the Virtual Museum for navigation and discovery within the collection is imported from the Australian Museum’s collection management system. The current collection management system is the Museum’s third generation of systems computerising its records of the Pacific collection.
To understand the evolution of the Pacific collection’s metadata we give an overview of the typical life cycle of their records. The Australian Museum acquired the objects in its Pacific collection from many sources over 150 years. The process of adding an object to the collection is reasonably uniform and easy to illustrate by example. The ‘fish hook’ was entered into the Australian Museum’s ‘Register of Ethnology’ on September 22, 1971. This entry is the first association of collection metadata with the object, and allocates its registration number. At some later point an index card was created which included the object’s provenance, and more detailed descriptive text, and (on its reverse-side) the object’s physical dimensions.
Later, as objects are added to the content management system, they are further described, and have a simple, practical corporate taxonomy applied to them. The spreadsheet documenting the Museum’s taxonomy presents the ‘organisational warrant’ for the metadata, the normalized way of describing things at the Museum. The Australian Museum’s “Archaeology and Anthropology” taxonomy is two-level, with 27 categories and 709 object types distributed across those categories. The development of the current taxonomy for the cultural and archaeological collections was developed by the Australian Museum’s Stan Florek, as an alternative to large general systems which tend to ‘lose’ objects in their many alternate paths.
From information collected during preparation of an initial 400 objects for the prototype of the Virtual Museum, we estimate that only about 45-50 percent of the objects in the collection have an entry in the electronic collection management system, a common problem with all large historical collections. Nearly all objects need metadata “scrubbing” to bring them up to a uniform exhibition standard.
Scrubbing involves normalising spelling and thesaurus checking, for instance testing whether “mother of pearl” or “pearl shell’ should be used or whether a “dagger” should be a tag or whether “knife” is preferred. The Virtual Museum project revealed that an average of one hour’s effort per object is required for basic metadata entry and scrubbing, and another hour to write an interpretive label (reminiscent of the descriptive card in a museum exhibition case).
So, while the metadata and photography adds enormously to the value of an object for research and Web-based exploration, there is a significant cost involved in establishing an adequate information base for it.
In reality there is very little ‘overhead’ in preparing objects for publishing using the current Virtual Museum infrastructure. Nearly all the museum’s effort goes directly to improving the documentation of the collection – which is its core business. Also, the fact that this improved information can be immediately published to the Web without a lot of intervention by technologists, rather than lying relatively hidden in the collection management system, is a great motivation for the museum staff.
In short, a key finding of the Virtual Museum of the Pacific is that the completeness of the metadata for the Pacific collection is enormously variable and therefore considerable work has to be done by the Museum and its staff beyond that performed by the Virtual Museum developers. We expect that this observation applies equally to other museum collection content and is an important consideration in developing Web-based Virtual Museums.