The Dictionary of Sydney provides an opportunity to explore leading edge technologies for encoding and presenting historical information.

The possibilities for delivering information to end-users change too rapidly for any one project to keep up with them all;  therefore, the emphasis is on designing and populating a flexible repository of information, encoded and cross-referenced in such a way that third parties can devise innovative and imaginative end-user applications – whether that is as a website, as mobile information, print-on-demand, a web service or as yet unthought-of possibilities.

The initial website we have built will become just one of many windows into an ever-growing information repository.

The base for technical work on the Dictionary is the Arts eResearch (AeR) at the University of Sydney.

Some of the many aspects of the technical research are described briefly below.

Open source

In general, the project works within an open-source environment. This means we attempt to identify open-source solutions to the Dictionary’s requirements (without excluding commercial solutions where they are appropriate and affordable) and it also means that any technical insights gained in the course of the project will be made freely available. These focus on information architecture, approaches to integration and the use of standards, less so on straight software development.

Markup and structured text

All the text and meta-data in the Dictionary is encoded in XML for easy markup and manipulation. Markup is both structural (tagging the text to identify elements like sections, paragraphs and sentences) and semantic (identifying links to related information such as linking a reference to a person to a more detailed entry on that person). We have used TEI as the markup standard for these purposes.

Knowledge engineering

The essence of history is that everything is connected and the categorisation of the Dictionary’s material is absolutely critical to delivering a richly contextualised experience – we avoid rigid hierarchies of categories. Technologies and standards associated with the semantic web are being investigated as ways to facilitate this as well as sharing and re-using controlled vocabularies.

We are closely investigating the use of RDF (Resource Description Framework) to define and manage descriptive metadata for the Dictionary’s resources. Models which define the use of RDF are described as ‘Ontologies’.

Concepts and ‘Entities’

In order to organise the Dictionary’s material – we have developed a model of Concepts (such as Transportation, Public Architecture, Theatre, Politics – an endless list of topics and areas of interest) and Entities  – as we call them – (Buildings, Structures, Events, People and Organisations). Categorisation models like these may seem at odds with Web 2.0 approaches where everyone is free to tag content in their own terms and develop folksonomies. Our approach in the Dictionary is to define the structures we need to deliver a rich experience but to keep those structures separate from the content so that it is possible to apply different categories (or ‘tags’) if appropriate. We are really trying to have the best of both worlds.

Digital repositories

We have investigated flexible digital repositories for storing digital objects long-term such as Fedora. Images and other multimedia are also essential components of the Dictionary. For initial identification and categorisation of multimedia resources we use Portfolio from Extensis. Essays and other text entries need to pass through an editorial workflow. To date, we have been using DRTrack, developed at Arts-eResearch. We also use the Open Journal System (OJS) to publish essays in a more traditional journal format. This also enables us to offer contributors an editorial process compliant with academic assessment standards.

Space and time

These are critical concepts for the Dictionary – we have created a framework of time and space into which the Entities referred to above can be placed. This will enable us to consistently show those objects in the context of, for example, maps and timelines.

We want to be able to do spatial queries on the Dictionary – for example, having identified a structure, say a bridge, on a map, the user should be able to ask to see other nearby bridges, other nearby bridges built by the same engineer, other nearby bridges built by the same engineer and built before 1885- and so on.

Other projects

We are constantly on the lookout for projects similar to the Dictionary which we can both learn from and increasingly contribute to. Some of these are listed below:

• Te Ara Encyclopedia of New Zealand
• Encyclopedia of Chicago
• MultimediaN N9C Eculture project
• National Science Digital Library
• Henry III Fine Rolls project
• Historical Event Markup Language (centered at the Dept. of Classics, Mount Allison University)
• Vicodi European Union history project
• Trove
•  Queensland Historical Atlas