One for all and all for one: towards a common information environment

Archaeologists don't always recognise the wide range of institutions in which we work. Some are familiar cliches of film and television: the university, the museum, the laboratory and the field. Others are familiar to the profession, but less obvious to the public: libraries, archives, national government agencies and local government planning departments. This represents a vast array of institutions: local, national and international; commercial, public and private; academic, amateur and professional. Each is to some extent dependent on information, and each depends to some extent on the others to preserve, produce, manage and distribute information for them. Though we seldom have cause to reflect on it, the practice of archaeology depends on information moving freely across many different sectors.

This is a microcosm of the predicament of electronic government. 'E-government' relies on large amounts of information moving seamlessly across and through a whole range of institutions. Above all, this information must be coherent to users.

This is not to say that e-government and the systems it requires are unattainable: a significant investment in the last few years has given many agencies large amounts of digital data. Rather it suggests that archaeology is well-placed to inform development.

This was the goal set for the ADS and HEIRNET in April when we were invited to develop a demonstrator of a "common information environment". Requirements for this project included the use of open standards, a desire to release information that would otherwise be hidden, and a focus on geography as a unifying classification. A commitment to 'design around the needs of the user' ensured that it would not be sufficient to present a single interface, but to design tools that were flexible enough for users to choose appropriate interfaces. Over and above this, a commitment to deliver the system in two months ensured that the project had to be planned carefully and executed promptly. The resulting demonstrator is a tribute to the staff and partners who contributed to its successful and timely completion.

Image of My HIERPORT.
Click on image for larger view.

'My HEIRPORT' - choosing different views for different purposes.

The resulting demonstrator - under the working title of "My HEIRPORT" - can be viewed in three different perspectives: the user, the information that it contains, and the technology used to support it. Perhaps most importantly it moves from search and retrieval technologies towards use and interaction.

Recognising that different users have different levels of skill and expectation, we identified six modes of user that may have an interest in the historic environment. A subtle but significant change in thinking took us away from having categories of user to identifying different types of behaviour that users adopt. Thus, rather than presenting a single interface, we developed six inter-related presentations. These interfaces provided search and retrieval tools, but located them firmly in the context of other aspects of the information environment: learning objects, community spaces and news services.

For those interested in exploring their local environment, we developed the "My Historic Environment" interface. We presumed that researchers had more background knowledge and a desire to get straight to information quickly, so developed a "Researching" interface. We created a "Learning" interface that provides a structured engagement, while a technically advanced interface was designed for occasions when users need to understand what's happening "under the bonnet". Related to this was an interface for collection management. A further interface was designed for those using location aware and mobile devices.

The data presented through these interfaces came from a wide range of institutions: Durham County Council, English Heritage, RCAHMS, the Portable Antiquities Scheme, SCRAN, Edinburgh University Library, the Archives Hub, the Fitzwilliam Museum, Cornucopia, HEIRNET and the ADS itself. A common metadata standard was applied to give the data a uniform index, and rather than having to find or invent a new classification scheme for this diverse set of information, the geographic properties of the information were exploited. Thus, users were able to ask 'what is there in this vicinity, and get back meaningful results.

Though the presentation provided six different interfaces, these were based on a single foundation. A combination of standards were used, including Z39.50, OAIPMH and RSS. Different target databases were presented using the most appropriate protocols. These were then presented to the web as a series of modules which were repeated or edited to suit the needs of each view.

Three novel areas of development tested recent additions to the range of standards and middleware available. Nick Ryan of the University of Kent at Canterbury developed a module that was based on location aware devices - using hand held computers that track global positioning satellites Because the device already knows where it is, this raises the prospect of never actually having to phrase a question more complicated than 'what is there around here?'. Staff at the RCAHMS took the problem of map awareness further by connecting the outputs of search returns to their Web-GIS server CANMAP. This shows the possibility of using web-based GIS in a widely distributed environment by exchanging data in XML formats. Another innovative tool was brought by staff at EDINA, the JISC data centre in Edinburgh University Library. They used a developmental geographic gazetteer service to transform grid references into place names, then used those placenames to search the Old and New Statistical Accounts of Scotland. These historic texts have no grid references and none have been introduced: the middleware gazetteer used here thus shows that map-based searching need not be restricted to map-based data sets.

Image of My HIERPORT.

Image of My HIERPORT.
Click on image for larger view.

Selecting which databases to search: two views of the same question for different types of use.

'My HEIRPORT' demonstrates what can be done: but it is not our intention to launch it as a service. Even so, many of the tools developed could be pressed into service with relative ease. Thus, as well as informing high level goals across different types of public sector agency, the demonstrator will feed into the continuing development of HEIRPORT, the Z39.50 portal that we launched in January 2002. It has shown that, with the right commitment from a dedicated group it is possible to create a common information environment across different public sector agencies. Archaeology, with its broad community in a myriad of different agencies is well placed to demonstrate what can be achieved.

William Kilbride
wgk1@york.ac.uk
For more on the Common Information Environment Working Group see http://www.jisc.ac.uk/index.cfm?name=wg_cie_home

In this issue ...