Mapping Information Resources

A report for HEIRNET by

David Baker
Gill Chitty
Julian Richards
Damian Robinson

Synopsis

Introduction

1 HEIRs (Historic Environment Information Resources) are developing in numbers and complexity, stimulated by rapidly advancing technology and selective funding opportunities, but with a lack of mutual awareness that is inhibiting useful co-operation and co-ordination. This perception prompted HEIRNET to commission a project mapping the landscape of information systems. Its brief was to identify all the key systems, their defined purposes and perceived roles and functions, in order to help reveal gaps and overlaps, and where existing information might be reused for other purposes such as education. It would also be the basis for a strategic vision of future development for HEIRs, enabling access and enhancing public benefit.

2 The first stage of this project devised an operational framework for understanding, conserving and explaining the historic environment, to encompass the processes of mapping and analysing complex issues. The model proposed covers the interaction between four elements, the historic environment in all its aspects, the conservation process in all its stages (including end-uses), the management of information, and organisational structures. (Appendix I - main report). How effectively these elements do, and do not, interact with each other can throw much light on current problems and help with devising future solutions.

3 Two obstacles ruled out the possibility of producing a precisely delineated map. More than twice the agreed coverage of about 50 systems was identified during investigation, and many more exist (Appendix 2 - main report). Except at the highest level of generality, and even then only provisionally, a diagrammatic map cannot effectively represent the multiplicity of linkages between different operational sectors and various uses of the historic environment. The principal ones are summarised in this synopsis.

4 The strategic vision required of this project envisages user needs at the centre of future information system development. It is desirable to maximise interconnectivity or interoperability between physically dispersed systems, moving away from earlier models of global information management. Provided this is done in the service of roles that relate clearly to user-needs, this will be more than just another technical exercise in restructuring. It has the potential to transform and strengthen the roles of the major organisations at national and local level, by defining them as points of access to both national and local data-sets, in the service of nation and community.

5 This strategic vision makes assumptions. All relevant activities must draw upon and contribute to appropriate information systems. Each existing system must be able to provide a clear statement of its scope, functions and purposes, as a means of communication and a practical aid to resource discovery. There must be a presumption, when new ones are created and existing ones re-engineered, of convergence towards standards that maximise the scope for interoperability between systems.

6 Its realisation depends upon three main elements.

(a) A central register of HEIRs, supported by the community of information systems, to make available details about their status and accessibility, in furtherance of government policies for cultural access, environmental conservation and education.
(b) A technical advisory facility to help projects use data standards and data structures that assist interoperability, and facilitate access by others wishing to bring research results to wider audiences.
(c) Concerted action by funding agencies to ensure that creators of information systems incorporate accessibility and interoperability.

The descriptive map

7 Holistic approaches to the historic environment have begun to be reflected in the information systems of organisations which have a role in managing the full range of environmental issues in their areas. Yet within historic cultural resources there are still largely unsolved problems arising from different approaches to archaeological recording and building conservation; also, these systems tend to be separate from those for museums collections and historical archives and records. There are similar gaps between the treatment of historic cultural and biological information resources. The integrative influence of Geographical Information Systems is a unifying factor in development generally, giving users locational points and scales for entry, but has yet to be taken up fully, especially locally, for the historic environment.

8 Issues of coverage and duplication emerging from the mapping process include:

(a) uneven and patchy coverage in some topic areas and aspects
(b) the rarity of consistent and stable single-subject corpora for defined aspects, with few HEIRs offering UK-wide coverage
(c) the lack of spatial data formats in local and national systems (especially for core designations) hindering compilation of integrated environmental management systems
(d) different attributes or levels of information about the same object recorded within several systems
(e) lack of the concept of shared core designation data to which management or specialist data can be added in other systems, leading to overlaps
(f) inconsistent use and non-standard holdings for bibliographic material
(g) low confidence in others data leading to repetitive collection, sometimes masked by variability in standards of information.

9 The conservation process is seen as a self-sustaining cycle of awareness, management and feedback into improved understanding, within a context of academic, cultural, economic, educational and social uses. The strategic centres for information are the national and local record holding bodies, NMRs, statutory datasets, SMRs and LEIMS ([generically] Local Environment Information Management Systems), together with the operational information systems of large historic estate managers. The coverage of information systems for conservation is related to organisational roles in environmental management and is most comprehensive in integrated systems that operate at the local scale. Some specialist systems draw upon, and theoretically feed back into, the local and national systems.

10 Issues relating to the conservation process include:

(a) little shared use of some national heritage conservation systems with local information systems, such as the computerised listed building record
( b) very limited access to the detailed data supporting published summaries of national heritage datasets, despite its relevance for local management
(c) unavailability of outputs from national reviews and assessments as a served digital resource for the conservation and research communities
(d) lack of an enhancement loop between specialist and generalist systems.

11 An understanding of the historic environment is the goal of academic research and the prerequisite of sound conservation practice; the conservation process is by far the largest producer of new information but the academic sector is a major user. Academic resources a re tending to provide the networks for co-ordination and resource discovery that the conservation sector has lacked. The many sources on which these networks draw are varied in location and quality.

12 Issues relating to academic usage include:

(a) weak information flows from the outputs of conservation action into accessible systems available for new research
(b) under-developed linkages between the knowledge derived from fresh academic enquiry and innovative conservation practice
(c) the potential for specialist research to improve data standards by developing technical vocabulary and thesauri
(d) the potential for a unified bibliographic / grey literature / field event index, of which only parts currently exist
(e) increasing distances between the knowledge areas of the historic environment making problems for the non-specialist in linking data from diverse resources.

13 Systems established in the 1980s and early 1990s with data-gathering initiatives primarily for environmental management require considerable redesign and reinvestment to fit them for the broader cultural and educational purposes that are now much more prominent in government policy. Museum-based systems have evolved more readily in this direction, and some national systems provide an entry point into resources available at that level. Such direct access to digital material is scarce or non-existent at regional and local levels.

14 Identified issues include:

(a) most national and local systems may not have been designed for these purposes but can usually supply information for others to use
(b) information initiatives for education to date have involved major investment; low-cost Internet resources are only starting to appear
(c) software for accessing data remotely or collecting it is often unfriendly (or non-existent), off-putting to voluntary, independent and education sectors
(d) system redesign does not always take these purposes into account
(e) uncoordinated funding strategies are not maximising potential synergies between national and local systems for these purposes.

15 The new orthodoxies of sustainable working have legitimised heritage, not just as a economic commodity for tourism and the leisure industry, but also as a central element in the social process of regeneration for places and communities. They have brought opportunities for adaptive re-use, adding value, new employment, recycling of material, revitalisation.

16 Identified issues include:

(a) assertions of IPR (intellectual property rights) and confidentiality can affect access to information: jointly controlled protocols are needed
(b) property managers have little incentive to make their information available more widely unless strongly involved in these other spheres
(c) undeveloped potential uses of information by commercial organisations in the property, development and tourism markets
(d) accessible and intelligible information for the local user is not available consistently in the right places or forms for general social projects.

17 Information management comprises a set of four inter-linked stages, creation, storage, dissemination and re-use. Whilst much data-creation is inevitably local, it can be done within regional, national or international contexts, and the other stages can be at any of these scales. Indeed, the stages are rarely combined in one organisation. Individual users with different needs for either direct or mediated information may pose questions at local, regional, national or international levels. Information about the historic environment is dispersed around the world, inside archives, libraries, and museums.

18 Issues arising from information management include:

(a) the lack of organised accessible information about information resources, effectively a barrier to resource discovery and to syntheses across regions, countries and time periods
(b) the design of existing information systems primarily to serve the internal functions of their organisations and not as information resources which can serve the wider information needs of research, education and other public and community interests
(c) integrated common access between information systems hampered by general problems of interoperability within and between systems at national and local levels and across specialist and general systems
(d) the existing network comprising bespoke connections, often with informal and unstructured links, resulting from organisations tending to develop information management strategies in isolation and unaware of the range of technical options
(e) the need for resources to develop and collectively adopt consistent data standards, especially for terminology control (and in older under-funded systems), in order to make progress towards interoperability
(f) the lack of consistent glossaries and mappings of professional terms to enable public access to HEIRs.

19 The organisational structures within which information managing systems exist ought to ensure that they have mechanisms for collection, maintenance and delivery, as well as appropriate contacts with other systems. Most systems can characterise their holdings, media and means of accession. In some cases of poorly managed un-indexed and non-digital record systems, intelligent access by external users depends upon staffed assistance which may be absent due to shortage of resources. Outside public information services, it is rare to find data structured and presented in a way which is designed to meet external user needs.

20 Issues relating to organisational structures include:

(a) few systems stating their purposes for collecting information and the range of potential users; very few collecting user numbers and types
(b) poor definition of organisational roles, resulting in overlap, inefficiency and lack of constructive joint-working with complementary systems
(c) weak organisational connections between interpretations and detailed underpinning information, together with significant costs of recasting information resources to make them more widely useful and accessible
(d) tensions between notions of organisational territories and ownership of markets to the emerging trend of shared information resources
(e) wide differences in the degree of quality assurance and provision of validated data between different individual organisations
(f) perceived differential requirements for access and retrieval between record holding and conservation organisations
(g) individual information-managing organisations developing their own future strategies with little awareness of other initiatives.

Case studies

21 Eight case studies illustrate the range of overlaps and other issues constraining the development of a strategic vision.

(a) A review of current initiatives in the UK Higher Education sector shows how the disciplines involved in the study of the historic environment are subject to broader trends in information management.
(b) Information systems for historic estates have to support management and research, and the needs of multiple user groups, within and outside the historic environment sector.
(c) Studies of historic gardens and landscape and historic religious buildings indicate sectors with a high risk of multiple initiatives and overlap.
(d) Studies of two chronological sub-disciplines, Roman Britain and the industrial heritage, also deal with multiple user groups and initiatives.
(e) Studies of artefactual research and scientific artefactual databases cover areas facing specific problems of integrating information and particularly needing the organising power of research themes.

National and local interfaces

22 These case studies indicate a transition in ideas, towards a distributed network of systems, its interoperability facilitated by convergent data standards and structures, and away from the impracticability of large monolithic systems and elaborately structured sets of smaller inter-dependent ones. Yet this will depend upon resolving problems at the organisational interfaces, between and within the national and local centres of the information landscape.

23 The dynamics and tensions of the national / local interface largely concerns relations between local SMRs / LEIMS and the older national records upon which many of them were initially based. Several instances of difficulties across the national-local interface were identified in the course of the survey. These included dispersing national or regional databases created by finite projects to local systems without provision for consistent overall maintenance, and the devising of projects at one level without adequate (or any) provision for benefiting the other. Similar difficulties can face specialised record systems and the management systems of the major historic property owners. Aerial photographs seem to be a particular problem.

24 The debate about functions seems to be moving from the relatively adversarial question of who should do what ? towards a more collaborative appreciation of distinctive complementary use-related roles, serving the nation and the local community respectively. At both levels new agendas of public access and social inclusiveness are leading discussion. Distinctive roles in research, conservation and explanation exist for nations and communities. Comprehensive multi-purpose information systems are needed to satisfy user-requirements at each level, managed and developed by professional staff with appropriately broad training. The latter can help reduce tensions over data-exchange and archival responsibilities through ensuring clarity about the concepts of, and responsibility for:

(a) curation, the primary archival duty of preserving, ordering, indexing and publicising
(b) holding, copies or summaries of original material curated at another level, on account of its significance for primary tasks
(c) accessing other systems through indexes held at either end.

25 Within the levels, communications between the major national systems in each country seem to be improving; the merger of English Heritage with RCHME and Wales ENDEX are two kinds of solutions at various stages of implementation. The national amenity societies need arrangements for shared information, especially on statutory designation descriptions. Local problems can arise between general systems variously located within County Councils, District Councils, Unitary Authorities, National Parks or joint arrangements between adjacent authorities. There are also weaknesses in communication between the different types of LEIMS (where they all exist) within environmental conservation, and between that area, museums and documentary archives. Detailed discussion of one county area (Bedfordshire) brings out many of the issues. How systems will be affected by any development of regional government is a further issue.

Information management for interoperability

26 National and local interfaces are a major, but not the only, set of relationships that needs to be regulated by good management of information. Inaccessibility is one of the principal barriers to effective synthesis. Networks are important facilitators of effective information flows. Technical solutions must be realistic, aware of disparate user-needs, and the costs and benefits of enabling access.

27 Creating information involves observation and recording, both of which are selective and purposeful. An explicit brief should always relate the purpose of the work to the relevant stages of the conservation process, and define the relevant levels of spatial interest within the historic environment. There should be careful selection of appropriate recording media for capturing data that varies in width of focus, levels of precision, and potential for manipulation.

28 Storage raises issues about effective structuring for multi-purpose retrieval, yet also depends upon the current state of information technology. With relational databases superseding flat-files, the scope for multi-variant analysis has greatly increased; the ability to link databases with GIS has greatly enhanced the scope for analysing the historic environment as a set of nested spatial dimensions. The event - monument - archive (EMA) data model highlights the need for working out consistent ways of storing and retrieving the complex sets of information that comprise interpreted hierarchies of monuments.

29 Dissemination and use involve making systems passively accessible to enquirers, and taking information, usually selected and interpreted, out to people. Many records are poorly equipped to serve the wider social purposes for which they were originally envisaged, before the invention of the World Wide Web. This is now the single most effective means for accessing, synthesising and publishing information on a global scale and is evolving towards machine-driven information retrieval, using communications protocols such as Z39.50. These enable a single user to make simultaneous queries of diverse, distributed, data resources; they require data creators to provide appropriately organised and digitised material only once. Such resource discovery metadata, or data about data, covers the nature of a body of information, its electronic location and the whereabouts of similar information. An emerging international inter-disciplinary standard is a metadata element set known as the Dublin Core. (Appendices 3, 4, 5, 6 - main report)

30 The technology now exists to allow all digital resources to be linked within a networked environment to enable cross domain discovery. This will revolutionise access to information for research, management and education. However, making information available is not the same as using it, and different users have different needs. Some can be directly satisfied through standard interrogation of information in appropriate formats; others, such as SCRAN, need to select information and re-present it before it can properly serve their purposes. There are perhaps three levels of remote access, to metadata that characterises types of data-field held, a metadata-based index to records, and the detailed information in those records.

Summary of Principal Recommendations

Strategy

1 Users needs should drive the future development of an agreed UK-wide developmental strategy for information systems serving the historic environment.

All HEIRs should be sensitive to user requirements; exploring them is a continuing requirement, not a one-off task. Users include future generations, whose specific needs cannot be precisely predicted, so must be bequeathed robust and flexible systems. The forthcoming CBA Publication User Needs Survey and the ADS survey of Digital Data User Needs (Condron et al 1999) provide snapshots of current user requirements, but will need to be refreshed in due course. Technological developments will make possible things not feasible before, and may bring to the fore previously unrecognised needs.

2 HEIRs should develop individually and collectively on a co-operative basis, working voluntarily towards mutual access and interoperability through co-operation inspired by awareness of identified roles and shared goals.

What HEIRNET represents is essential to this approach. It has a continuing role as a forum for the dissemination of information about current initiatives, especially through its web site. It is uniquely placed to encourage convergence and develop interoperability between HEIRs. It should endorse and refine the metadata standards for the UK historic environment.

Until the technical vision of full interoperability advanced in Section 5 becomes a practical reality there will be a continuing need for sensible definition of the responsibilities of the stakeholders as discussed in Section 4, especially 4.8.

Implementation

3 A central Internet register of HEIRs, supported by the community of information providers, should make available details about the status and accessibility of information systems.

Such a register would help users identify potential resources, and identify gaps and duplication in the coverage of HEIRs. It has precedents in several systems-about-systems encountered during the mapping exercise. It could also provide an optional clearing-house for communication over access and outreach projects, thus maximising opportunities to learn from developments and experiments by others. Registration might be voluntary, but made a condition of grant where applicable. Non-conformity with the MIDAS standard would not be a bar, but conformity with it would be specifically identified. Its success might be measured by the rate at which the register of information systems developed, and the amount of use made of it measured as Internet `hits'.

Descriptions of content and contact points would be a major step towards opening up access to information. The collected data summarised in Appendix II could be presented in a standard database format. Existing metadata standards provide a good means of indexing HEIRs for the purposes of resource discovery. Such index records should be made available for on-line searching, and the ADS catalogue ArchSearch provides one obvious entry point. Entries would ideally provide metadata and keys for inter-operable access, documenting Z39.50 protocols in use and indicating whether an HEIR is a Z39.50 target. Appendix V contains a Dublin core exemplar.

The central register should be owned by the community of information providers rather than a single government department or an NGO. It might supersede the EH / NMRs MIDAS Inventory Registration Scheme and the informal register of specialist systems which NMR has recently begun to compile. It should be, and be seen as, relating to all aspects of the historic environment, and not just primarily the concern of archaeologists. It will need maintenance and updating, especially validation of the data collected. There may several options for running such a register, combining the right skills and sending the appropriate messages. One option familiar to the consultants is a partnership of the Archaeology Data Service at the University of York, with its experience in the fields of metadata and interoperability, and the Council for British Archaeology, but acting explicitly as a lead body for the statutory national amenity societies.

4 A technical advisory facility should be made available in order to
* promote the use of data standards and structures that assist interoperability
* document and monitor the use of appropriate metadata standards and protocols
* advise on the production of high quality information and how to bring it to wider audiences through the proper use of information systems.

Consideration should be given to whether the technical advisory facility should be based on a single body, or distributed as a networked consortium of leading bodies in the four home countries. Whatever course is followed, it will need to build upon the service in documenting standards for use in relation to the historic environment provided by the NMR Data Standards Unit and the mda.

The technical facility should continue the documentation of appropriate schemes to develop communication between those bodies using the Dublin Core for the description of their HEIRs, and to monitor its application and effectiveness. It should facilitate the agreement of a communications protocol and profile mappings. It should discourage the proliferation of additional protocols, and should work to facilitate interoperability between potential partners. There is a need for substantial work on applications of Z39.50 and on the development of gateway interfaces. Gateways should be targeted towards their user groups and should develop interfaces appropriate to the level of prior knowledge to be assumed. Amongst HEIRs surveyed there are a number of existing users of Z39.50 protocols (ADS, SCRAN, Portable Antiquities) and others are interested in developing such systems.

A related task is the development and implementation of agreed UK-wide standards for terminology control, in the form of an adequate integrated subject thesaurus for British archaeology, as recognised by BIAB. The FISHEN group provides an appropriate forum but its remit and representation should be extended to the UK as a whole (although it must also recognise that the borders of interoperability do not stop there).

5 The issues of strategic roles and relationships in, and between, UK-wide, national, local and thematic information centres should be jointly addressed through the creation of strategic discussion groups of those involved.

For detailed discussion of issues of coverage, roles and relationships, see Chapters 2, 3 and 4 in the full report.

Resources

6 The technical advisory facility should pro-actively seek special funding to achieve critical steps in the developmental strategies for HEIRs.

7 HEIRs should normally develop and be maintained by adequate levels of funding according to the uses defined for them.

8 Agencies funding projects whose activities use or develop information systems should ensure that those projects conform to prevailing procedures and standards.

EH, CADW and HS can each play an important lead role here; ALGAO and ARIA have an important role in regulating the commercial sector. The AHRB and AHDS have a key role in promoting recording standards within the HE sector.