The ARCHES Project

by Mark Childs
University of Warwick, UK

The ARCHES project (Antiquity Related Collections Harnessed for Educational Scenarios) places photographs, images and animations of 3D computer models, and real-time navigable computer models from previous projects within a public online repository, available for teaching and research, and for use by students in creating essays, presentations and performances. The digital content is drawn from previous projects including the THEATRON Project and private collections of lecturers, including Dr.Zahra Newby from the University of Warwick's Classics Department. The first version of ARCHES is now available online through Didaskalia.

The aim is that people using an ordinary web browser will be be able to:

We also wanted the collections and the assets used within them to remain hyperlinked. This would enable anyone discovering an asset to see all the collections in which it had been used, or alternatively, anyone viewing a collection would be able to click on a specific asset and open up more information. The idea was that information could be found systematically through structured searches, or serendipitously by following the trail of links from collection to asset to other collections and so on. This function was thought to be a way to enable students and researchers to uncover material from other subject disciplines with which they might be unfamiliar.

The primary development phase of the project, which is based at the University of Warwick, ran from January 2003 to February 2005. It was a collaboration between the University of Warwick's School of Theatre Studies, the Classics Department and the Centre for Academic Practice. It was funded by the Joint Information Systems Committee (JISC) under the Exchange for Learning Programme (X4L). The project was directed by Professor Richard Beacham and Dr Hugh Denard in the School of Theatre Studies and managed by Dr Jay Dempster at the Centre for Academic Practice. The co-ordination and evaluation of the project was conducted by Mark Childs. We hope that ARCHES will continue to be developed in collaboaration with the Centre for Computing in Humanities at King's College London and other partners. What follows is an account of the first phase of the project, which is now complete.

It became apparent that there was no existing digital repository that could provide the functionality we needed. The University of East Anglia's Artworld system came very close, and so the software developers who designed that were commissioned to create a similar system for ARCHES.

Design considerations included how to make the assets and the collections searchable. To achieve this they were tagged with metadata (cataloguing information). The collections also were required to be deposited within JISC's national repository (Jorum), and so the metadata we created had to be compatible with their standards. The metadata specification had to be capable of describing, within classics and art-historical contexts, digital, visual and textual assets and collections of assets. No single existing specification answered to all of these requirements, but it was of paramount importance that our specification should be compatible with international metadata standards that are designed to enable online resources to be shared across projects, institutions and countries. Our specification is therefore a composite of elements from a number of recognised schemas: Dublin Core; Visual Resources Association; the Getty's Art and Architecture Thesaurus; and the UK Learning Object Metadata

The aims of the project were therefore constantly changing. Although at the start, the project's aims were entirely to do with implementing changes in teaching practice, once the project began these were substituted by considerations of a more technical nature, chiefly: how to create the repository. Creating the repository required us to resolve the problems of information management, i.e. what subject headings (taxonomy) we would use to arrange the assets; which metadata sets we should draw upon; how quickly and economically to generate the metadata required. These issues were all resolved, and the project participants have been able to contribute to the national debate on creating metadata, providing advice and support for academics within and beyond the University of Warwick on these areas; a field of expertise we had not anticipated acquiring at the start of ARCHES.

In March 2004, this system was sufficiently advanced to be used in the classroom. Dr Newby ran a session with four groups of students who used the assets stored within ARCHES to create collections, which they then presented to the rest of the class. The results were highly successful. The students reported that the software was easy to use, and that the learning activity had been a valuable one. The students' approval of the ARCHES system was underlined by their request that the resources for more of their modules be added.

A failure in the system at this time was the inability to use the subject headings effectively to create a working “browse” function. We were concerned from the outset that simply to provide a free text search for keywords would be problematic, since users would need to be aware of the range of resources available within the repository in advance of any search in order to use the repository effectively. We therefore drew the subject headings from the 'Style/Period' classification of the Getty Art and Architecture Thesaurus, in which styles and periods are arranged hierarchically according to an internationally-recognised schema. This should have meant that a student could browse through, for example, artworks from the Greek Classical period (750-490 BCE): the repository would automatically have displayed all assets so defined, and also assets within subsidiary parts of the hierarchy, e.g. Early Classical (480-450 BCE), High Classical (450-400 BCE) and Late Classical (400-330 BCE). In the end, however, it proved impossible for the developers to achieve this. As a short-term compensatory measure, we have replaced the free text search with a drop-down box of keywords used in creating the metadata.

A further remaining issue with the ARCHES system is the overall look of the repository, and in particular its Didaskalia interface. We hope to be able to find the resources to commission a graphic designer to produce a more attractive interface. However, the resulting repository is both an effective demonstration of the proof of concept, (that students can build learning objects easily and that the two-way linking of assets and learning objects works as a means to discover resources) and a workable tool for learning and teaching activities.

In addition to the direct development of ARCHES software, it is possible that certain functionality developed in ARCHES may also be integrated into other repositories such as Edinburgh's SCRAN archive and JISC's Jorum. We also hope that the models of good practice developed through the ARCHES project, those of creating metadata effectively and economically, and identifying appropriate methods for describing collection metadata, will continue to be drawn upon by colleagues within the HE sector.

For the time being at least, ARCHES will continue to be maintained by IT services at Warwick, and its resources will therefore be available for all lecturers, researchers and students who wish to view them and use them in creating their own collections. Anyone wishing to add to the resource, either by adding new assets, or by annotating the assets already in there, is welcome to do so, and will be able to do so for as long as the system is supported. Contact Mark Childs at the Centre for Academic Practice for details.

Mark Childs

Mark Childs is an Educational Developer in the Centre for Acacdemic Practice at the University of Warwick.