Search   |   Back Issues   |   Author Index   |   Title Index   |   Contents

Articles

spacer

D-Lib Magazine
June 2005

Volume 11 Number 6

ISSN 1082-9873

Building Interoperability for United Kingdom Historic Environment Information Resources

 

Edmund Lee
Standards and Guidelines Manager
English Heritage National Monuments Record Centre
Kemble Drive
Swindon, SN2 2GZ.
United Kingdom
<edmund.lee@english-heritage.org.uk>

Red Line

spacer

The Historic Environment - an Information Landscape

Over the last ten years in the United Kingdom cultural heritage profession, the term 'historic environment' has come into use as a deliberately all-embracing concept. The historic environment is all around us. It consists of the totality of those aspects of the built heritage, archaeology, and current and past landscapes that form both the subject of study for academics and a 'sense of place' for those who live and work within it. This holistic approach presents particular challenges to the designers and managers of Historic Environment Information Resources (HEIRs), the information systems used by both professionals and enthusiasts for documenting and analysing the historic environment. This article introduces the approach developed within the United Kingdom heritage sector to tackle these challenges.

The historic environment is distinct from the concerns of the more familiar cultural memory institutions – the museum, the archive and the library. The distinctions can be illustrated as follows:

First of all, for any given place or feature of the historic environment, there are often many organisations that have an interest in it. This contrasts with a museum object, which will generally be documented by a single curating authority. A feature of the historic environment – say a prehistoric burial mound – may be recorded simultaneously by any or all of the following HEIRs: a local authority for development control purposes, a national body for purposes of legal protection, the landowner for land management, or a thematic national survey of sites of a particular type.

Likewise, places within the historic environment do not carry information inherent within them about their nature and subject matter. The data need to be derived by research, such as comparison with similar places or features (for example, a survey of burial mounds). There is a distinction, therefore, between an HEIR and an archive. Archive items (photos, documents, plans, will most frequently contain inherent information and context – for example, within a collection – to allow them to be documented appropriately.

Finally, no two features of the historic environment are identical. Indeed it is most often the very diversity of features that is the subject of interest. Two prehistoric burial mounds recorded by an HEIR cannot be treated in the same way as two copies of the same book in a library. A standard, publicly available catalogue description cannot be applied to both.

The Interoperability Challenge

A wide range of HEIRs has developed independently within the United Kingdom [1]. This presents a major problem of interoperability. To achieve a full picture of the existing knowledge of a single feature in the historic environment, it is often necessary to draw upon information from a variety of different HEIRs. To review the full range of knowledge about a particular aspect of the historic environment on a national scale, it is necessary to consult dozens of different HEIRs.

However, in most cases these HEIRs have different physical designs (that is, relational table structures), and, quite often, have different underlying logical models. (Note this is therefore a more complex issue than simply dealing with differing file types such as .txt, .mdb, .xls, etc.). Various HEIRs will support different types of search and thus, will not provide consistent output. They may use different indexing vocabularies (or worse, none at all). As a result, the process of transmission of data between these incompatible data structures is complex and costly. As illustrated by the diagram in Figure 1, each attempt to transfer data between systems, or to make data available via the Internet, requires the design of a different tool. Consequently, providing data to multiple partners quickly becomes prohibitively expensive. The differences in HEIRs mean that any given 'tool' is unlikely to work for transferring data in both directions; thus, active collaboration or 'round tripping' of data doubles the effort required. The routines developed are also vulnerable to any changes in technology by either partner in the exchange. Resolution of this problem in the UK historic environment sector has taken a twin track approach. Technical development is necessary, but first there is a need for the diverse organisations to commit to working together to embed their technical solutions in real-world working practice.

An image showing the current situation at FISH

Figure 1. A complex set of different information management 'tools' is required for an HEIR to provide data to a variety of external partners.

Nurturing an 'Information Ecology' for the Historic Environment

The phrase 'information ecology' was first adopted by Nardi and O'Day [2]. Their study describes the need for an information system to be regarded as one part in an interdependent community that must support the needs not just of the system builders, but also of the system's end users. The term ecology is deliberately chosen by Nardi and O'Day to evoke the sense of dynamism and fragility inherent in a biological system. The same approach can, I believe, be applied to the development of the technical solutions that underpin developing HEIR interoperability. The ecology metaphor emphasises that information systems and data standards can only succeed where they also relate to the needs and experience of all parties involved. As in a biological community, no one organisation can predominate to the exclusion of others without an ensuing catastrophe.

Two issues serve to illustrate this point: intellectual property rights, and access to confidential or sensitive data (such as the exact location of sites that have yielded valuable artefacts). No technology to support interoperability of data will succeed without attention to issues such as these that might otherwise constrain or prevent the movement of information between HEIRs.

In the U.K. historic environment sector, beginning in the early 1990s, such an ecology has been evolving and resulted in the establishment of FISH, the Forum on Information Standards in Heritage in 2000 [3]. This 'ecology' seeks to co-ordinate the activities of three different groups of 'organisms' (as shown in Figure 2):

  • The Heritage sector. This sector presents the requirement for information standards, driven by evolving needs to record an increasing variety of heritage subject matter and to provide access to the data.
  • Standards Professionals. These professionals respond to the Heritage sector, providing standards that meet that sector's needs, but they also develop 'standards for standards' (for example, the development of the BSI standard for thesaurus construction [4]).
  • Software developers. These developers implement the standards as part of their service to the Heritage sector, but their work is also driven by advances in operating systems, software, hardware and communications infrastructure.

An image that shows the components of the FISH 'information ecology'

Figure 2. The FISH Information Ecology.

The purpose of FISH is to harness together the interests of these sectors to provide the best possible service to those who have an interest in the historic environment or cultural heritage in its broadest sense. The dynamism – the flow of energy round the ecology, as it were – stems from the pattern of consultation and collaborative working that has emerged within FISH.

At the heart of FISH is an email discussion list [5] with some 400 members from across the U.K. and other countries worldwide. The discussion list is used to air ideas, share information, and seek advice and comment. The starting point for work on a new area of standards development is often a structured discussion or 'e-conference' held on the FISH list [6]. When a new standard is developed to a draft stage, the list can be used to contact potential reviewers to ensure that the new standard is relevant to the needs of the whole 'ecology' [7]. A formalised peer review then either supports the approval of the new standard by the FISH steering committee or recommends further re-working of the standard. Additional structure and robustness is given to the process by the adoption for FISH projects of a formal project management methodology, PRINCE2 [8], a system widely used in UK public sector information technology projects. Together these measures are intended to develop and maintain a collaborative and dependable approach to shared work. The current focus of attention is on the FISH Interoperability Toolkit.

The FISH Interoperability Toolkit

The work of FISH and its predecessors has done much to develop a degree of commonality of content and terminology between HEIRs. The key documents are MIDAS [9], published in 1998, and the INSCRIPTION terminology suite [10]. MIDAS, currently in its 3rd reprint, defines the concepts or 'units of information' (basically metadata elements) that are recommended for inclusion in various types of record of the historic environment. INSCRIPTION provides access to published terminology standards suitable for use by HEIRs. However, the costs of export, manipulation and migration of data between systems are still prohibitive. To tackle this issue, FISH envisioned an interoperability 'toolkit', a range of protocols, formats, agreements and training materials necessary to provide HEIR developers and managers with the means to move data between systems [11]. A contractor, Oxford ArchDigital [12], developed the toolkit, with funding supplied by English Heritage and the National Trust. The toolkit initially has three technical components: MIDAS XML, the Data Validator Tool, and the Historic Environment Exchange Protocol.

Image showing the workflow for the FISH Interoperability Toolkit

Figure 3. The FISH Interoperability Toolkit sits between different HEIRs to assist with transfer of information.

MIDAS XML

MIDAS XML is a set of World Wide Web Consortium compliant Extensible Markup Language (XML) schemas, based upon the MIDAS data standard. The separate schemas cover areas such as Monuments, Events (i.e., investigations), People, Spatial Data, etc. The design is deliberately loose so as to include the wide range of current HEIRs. The design has also been informed by other standards, specifically the ontological thinking incorporated in the CIDOC Conceptual Reference Model (ISO 21127) [13]. This will ensure that the basic schema can be extended to meet the wider range of information envisaged by the forthcoming second edition of MIDAS, and also forms the start of links to other cultural memory institutions covered by the Conceptual Reference Model domain. Incorporation of appropriate elements of the Open GIS Geography Mark-up Language (GML) standard [14] and the Well Known Text (WKT) representation of spatial data will ensure compatibility with well-established Geographic Information Systems that are widely used within the historic environment sector.

All other parts of the Toolkit rely on MIDAS XML. It forms, as it were, the 'handle' to which other tools can be connected.

The Data Validator Tool

The Data Validator Tool (DVT) is an application developed to validate the content of MIDAS XML files. This tool will check the content (i.e., presence or absence) of the elements in MIDAS XML data against defined standards. It also matches terms used in the XML file with the terminology standards maintained within the INSCRIPTION framework and reports on possible problems. The DVT is flexibly designed so that a database of relevant rules and standards can be maintained to allow for the evolution of the Validation procedures as experience develops.

Together MIDAS XML and the Data Validator Tool promote the creation of standardised information resources that can be passed from organisation to organisation, or system to system, or deposited in digital archives (as shown in Figure 3). They also form an infrastructure upon which other tools and services can be developed.

The Historic Environment Exchange Protocol

The Historic Environment Exchange Protocol (HEEP) is a Web Services protocol for the remote querying and retrieval of data from MIDAS XML compliant HEIRs. Protocol requests will be sent using specially developed XML schemas, with the results delivered in MIDAS XML. This exchange protocol is the most exciting and innovative tool in the FISH Interoperability Toolkit. It will support machine-to-machine exchange of 'live' information between HEIRs, as opposed to the manual approach of record copying, export and import of data. While MIDAS XML schema dictate the structure used to exchange data between systems, the HEEP provides complaint computers with instructions for secure and structured transmission and data exchange.

Progress and Prospects

Technical development of these three tools was completed in December 2004. The Forum on Information Standards will maintain and develop the FISH Interoperability Toolkit on behalf of the historic environment sector. Future development may tackle other impediments to interoperability. One example is the problem of concordance between records derived from different HEIRs where it is important for one system not to include duplicate records for the same place. It is hoped that using a standard format for the data will support the development of software tools to automate the comparison of different datasets to identify sites recorded in both.

Further developments will be discussed via the FISH discussion list and promoted via the FISH website. All who are interested in these developments are welcome to participate.

References

1. Baker, D. Chitty, G. Richards, J. Robinson, D., HEIRNET Mapping Information Resources. Council for British Archaeology, York (1999) <http://www.britarch.ac.uk/HEIRNET/rep1.html>. [Accessed June 2005]

2. Nardi, B., O'Day, V., Information ecologies - using technology with heart. Cambridge, Mass. MIT Press London (1999).

3. Forum on Information Standards in Heritage website <http://www.fish-forum.info>. [Accessed June 2005]

4. British Standards Institution, British standard guide to establishment and development of monolingual thesauri / - 1st rev. - London. BSI, (1987) - (BS5723:1987) (ISO2788-1986) [currently undegoing revision]

5. Forum on Information Standards in Heritage discussion list. Available from <http://www.jiscmail.ac.uk/lists/>. [Accessed June 2005]

6. Lee, E., HDM Information Sheet Is an e-conference right for you? English Heritage Heritage Data Management Swindon (2001) Available from <http://www.jiscmail.ac.uk/files/FISH/Is_an_e-conference_right_for_you.pdf>. [Accessed June 2005]

7. Lee, E., Bell, A., HDM Information Sheet How to do a Peer Review. English Heritage Heritage Data Management Swindon (2001) Available from <http://www.jiscmail.ac.uk/files/FISH/Information_Sheet_How_to_undertake_a_peer_review.pdf>. [Accessed June 2005]

8. Managing Successful Projects with PRINCE 2. Office of Government Commerce, 3rd edition, The Stationery Office London (2002). Details are available online at <http://www.ogc.gov.uk/prince2/>. [Accessed June 2005]

9. Lee, E., (ed), MIDAS - A Manual and Data Standard for Monument Inventories. Royal Commission on Historical Monuments of England. Swindon (1998). [3rd reprint available via FISH <http://www.fish-forum.info>.]

10. FISH, INSCRIPTION. Available via <http://www.fish-forum.info/inscript.htm>. [Accessed June 2005]

11. Bell, T. and Lee, E (eds), FISH Interoperability Toolkit, version 1.0 (2005). Available via <http://www.fish-forum.info>. [Accessed June 2005]

12. Oxford ArchDigital website, <http://oxarchdigital.com>. [Accessed June 2005]

13. CIDOC Conceptual Reference Model website, <http://cidoc.ics.forth.gr/>. [Accessed June 2005]

14. Open Geospatial Consortium: (1999 - 2004) GML – the Geography Markup Language. Available via <http://opengis.net/gml/>. [Accessed June 2005]

Copyright © 2005 Edmund Lee
spacer
spacer

Top | Contents
Search | Author Index | Title Index | Back Issues
Previous Article | In Brief
Home | E-mail the Editor

spacer
spacer

D-Lib Magazine Access Terms and Conditions

doi:10.1045/june2005-lee