Information Studies Research

Digital Collections

7.1 Consultations and research on behalf of the HLF demonstrated that the retroconversion of catalogues and the creation of core data sets (e.g. biological databases, sites and monuments records in archaeology, the cataloguing of uncatalogued holdings) were perceived to be fundamental to all future uses of ICT in the heritage sector. They provide digital material to form the backbone which supports more innovative heritage uses of ICT, such as multimedia applications and Websites. As organisations digitise their holdings for particular purposes they must consider that there will be numerous other ways in which these materials might be reused. Digital content can be manipulated, sorted, merged with other materials, analysed, and reconstituted to meet a wide range of heritage needs. In museums, libraries, archives, biological recording centres and archaeology, the effective use of ICT depends upon the existence of good documentation systems. By giving priority to the conversion of existing paper and other records to a digital format, and by improving the character, quality and security of existing databases and making them more widely available, institutions create assets with long term economic and social value. Recent projects that have been coming forward to the HLF provide evidence that organisations realise that once catalogues, finding aids and other materials have been put into digital form, they can form the basis of information services for management, to inform visitors, as part of outreach services, and for curatorial purposes. These resources are the reusable raw materials of the information society.

7.2 Simple transfer of existing catalogues to digital form could be excluded from HLF funding, unless it results in the creation of a genuine value-added element to public appreciation and understanding. In some cases, simple retrospective conversion of a catalogue might be justified because it would support a better understanding of a particular collection and, as a result, improve its conservation. It does this by providing librarians and archivists with the raw materials they need to identify particular items or groups of holdings meeting certain criteria (e.g. books printed before a particular date, items printed by a particular publisher). In that event it must be realised that the process of transfer from one form of catalogue storage to a digital format may be both time-consuming and expensive, without necessarily being productive in the sense of increasing the value of the information. Here, added value could only be achieved if it subsequently became possible to use and manipulate the data more efficiently in new ways to promote either greater understanding, or increased accessibility of information to a wider public and range of users. One of the many heritage projects which has already used ICT to improve public understanding is the Canadian Heritage Information Network (CHIN); this provides online the National Inventories of Canadian Museum Collections, covering close to 25 million objects.

Recommendations
 
4. The HLF should support the creation of digital resources through collection of new data.
 
5. The HLF should support the retroconversion or digitisation of material where value is added to the data during the process of conversion, whether this is a by-product of its transfer to digital form or as a direct result of enrichment during transfer.

7.3 Much of the content created in digital form is raw material such as biological and habitat observations, entries in Sites and Monuments Records (SMRs), and collection documentation. Heritage professionals argue that a great deal of this material cannot easily be used by the general public without a layer of interpretation, whether by means of textual descriptors or software filters. This implies a need to select the information to be made publicly available and to decide how this selection will be extracted and presented from, say, a database, even though the latter may incorporate far more data than the end product. For example, most SMRs normally store raw data that are primarily intended for use by professionals and not by the general public. If they included some explanation or description, as well as images, and could be accessed by easy-to-use software, their public accessibility would be dramatically improved. The claim is that it is difficult for children and non-specialists to use them, without some kind of interpretation. The professional users who know the data best, who created or manage them, should ideally be the group to provide an interpretative layer. This would be an expensive project and would need to be continually updated as the interpretations of the material changed. Funds for this kind of work would not come from the holding bodies, such as the planning units of local authorities where many of these SMRs are located, since their users do not need this level of interpretation. Research and experimentation in this area is necessary before HLF starts funding implementation. The possibilities include software filters, intelligent databases, and collaborative work between museum education officers, communication specialists, and data creators and providers.

Recommendation
 
6. The HLF should expect heritage organisations to demonstrate in their applications how they propose carrying out the generation of high-level synthesis to make material more understandable.

7.4 In general, there are three main types of digital resources: products resulting from the retrospective conversion (retroconversion) of analogue material into digital form; resources created solely in digital form; and resources created in digital form as a by-product of administrative or commercial activity. The first two types are of current concern to the HLF. The third may eventually be a heritage asset, but its creation and preservation are currently not likely to be a pressing issue for the HLF.

Digital Collections

7.5 New digital resources can have heritage value. A great deal more information is needed about the cultural and natural heritage to preserve, manage and enjoy it, and to monitor both past projects and its current state. The UK has a long tradition of involvement with its natural heritage but if we want to continue this, we need access to the knowledge and standards for planning. There is a need for collation to be done by a small number of recognised organisations. Collection of data to create databases of natural heritage (e.g. animals, habitats, plants), the built heritage, or sculpture and monuments, is fundamental to the development of resources which contribute to the curation and enjoyment of the heritage. Much of the UK’s heritage is either not recorded in detail or is not recorded to a consistent standard. The benefits that have accrued from use of heritage records so far created provide the justification for further efforts to build more heritage databases.

7.6 Britain’s heritage assets and collections are not all held in public sector institutions (e.g. museums, libraries). Some—monuments and sculptures, for instance—rest in open spaces such as parks. Many historic automobiles, airplanes, and ships remain in private hands, or are held by trusts or other organisations which care for them as a public good. The National Historic Ships Committee (NHSC) database (already described in ¶ 1.1), provides an excellent example of the value of newly created data resources to the preservation and understanding of the heritage. Although only partially complete (900 of the 3,000 ships have been entered), it has already demonstrated its usefulness as an authoritative source when assessing the heritage merit of historic ship projects. The success of this project suggests that similar resource creation projects will bring benefits for other heritage assets.

Recommendations
 
7. The HLF should support the use of ICT to create new digital resources of heritage merit, and among projects proposing to do this kind of work a priority should be given to projects which involve community effort in the collection and collation of the data (see ¶s 9.11-9.14 on collaboration).
 
8. The HLF should not be nervous about supporting work involving creation of digital records of heritage assets which are not in public conservation management where these materials will encourage and inform conservation, assist public understanding, and provide a record of objects at risk of loss.

Retroconversion

7.7 Retroconversion is the process of converting analogue resources to digital form by representing the information they contain as sequences of characters and numbers (by comparison, see ‘digitisation’ in the next section). The retroconversion of library catalogues, archival finding aids, habitat and natural records, and museum collection documentation from printed, typed, or manuscript catalogues to digital form is an urgent priority. Once these resources are converted they will:

• provide heritage organisations with better knowledge about their holdings;
• permit organisations to develop coherent strategies to better manage and care for their collections;
• make it possible for institutions to develop a picture as to how their holdings relate to the holdings of other institutions;
• form the framework for programmes of digitisation by providing suitable databases for linking the information about holdings and digital facsimiles; and
• form the framework for public outreach by making it possible to give general access to information about the holdings and to allow institutions to better identify material of community interest.

7.8 Libraries are far ahead of most heritage institutions in the retroconversion of inventories of their holdings. For example, most archives are behind libraries in the retrospective conversion of lists, finding aids, and catalogues to digital form. The Public Record Office’s Archives Direct 2001 programme provides an example of the impact that retroconverted materials might have on archives. At the core of these activities is a project to provide online access to the descriptions of the PRO’s holdings, including the information currently held in lists and the Guide. When completed, the retroconverted finding aid to the 8 million records in the PRO’s care will provide the framework from which other services can be developed and managed: 24 hour online access to information; improved and more comprehensive catalogues; more flexible ways of using the finding aids; interactive document ordering systems; and online access to digital images of some records. This project demonstrates the benefits of technical planning and of the decisions to use open standards and to convert data into a portable format. As a result of these latter choices, these data can be searched and presented in a wide range of ways, and transferred to new technological environments as necessary.

7.9 The remarkable quality of archives and special library collections in the UK is widely acknowledged, although many of the resources remain unknown to the general public. There are millions of catalogue records for material held in special collections awaiting conversion, along with catalogue entries for older collections. These exclude catalogue records for more than 9 million non-print items (e.g. archives, manuscripts, photographs, slides, films, sound recordings, etc.) which need to be converted to digital form, and more than 12 million non-print items which have never been catalogued.

7.10 The Royal Commission for Historical Monuments (England) (RCHM(E)) and its partners, English Heritage and the Department of National Heritage (DNH), converted from paper to digital form and indexed the statutory lists of historic buildings to a consistent national standard. Once converted, it was possible to search, compare, and extract information within the records in ways that had never been feasible with the old printed records. For example, it is now feasible to search all buildings records for particular architectural features or materials, or even to identify the location of all structures of particular kinds, with greater speed and precision than was possible with the printed resource.

7.11 Retroconversion of materials from the biological sphere is also critical. For example, the Countryside Council for Wales has habitat survey information collected over a number of years which gives a complete overview of the habitats of Wales. In digital form it would be possible to create a basic foundation for more detailed planning and use. Digital maps can be created showing the distribution of particular habitats and these can be studied with other evidence such as maps showing species distributions.

Recommendations
 
9. The HLF should support the retroconversion of catalogues, inventories, and other analogue information resources used by the heritage sector where the digital version will bring new benefits to the public sector and where the resource has long term value.
 
10. The HLF should positively support the use of ICT to permit archival institutions to retroconvert their finding aids to digital form.
 
11. The HLF should only support those projects that demonstrate that during retroconversion every effort will be made to add value to the resources being converted.
 
12. The HLF should only support digitisation programmes of documents or images at archives which have retroconverted, or are in the process of retroconverting their finding aids.
 
13. The HLF should support the retroconversion of habitat and biological records, and archaeological archives.

Digital Imaging

7.12 Digitisation is the process of creating a binary representation of an object—a photographic print, manuscript page, or text document—which can be manipulated, stored, transmitted, and displayed using digital technologies, such as a computer. Digitisation is an expensive and time-consuming process in which attention must be paid not only to the processes of selection and capture, but also to those of description, storage, distribution, and long term retention. The justifications include:

• good conservation reasons for creating digital copies of decaying or heavily used resources. Digital copies make it possible to reduce the wear-and-tear on the originals. For example, some materials have decayed so badly that, if handled, they will be lost altogether and in this case the substitute value of a digital version is its own justification. In other cases the rarity of the object makes it impossible to give access to every individual who might wish to use it, but the digital copy does not suffer this limitation.
• improved access—digitisation brings a major advantage to potential users in that materials can be easily distributed across networks and, unlike a conventionally printed resource, there can appear to be as many copies as there are users. In an educational environment, or for rural communities, access to digital libraries means that local libraries need not use scarce resources to maintain collections that are used by few or are in very high demand for only short periods and would need to be retained in multiple copies to meet that need. Digital versions of certain unique materials, such as the British Library’s Beowulf manuscript or National Library of Scotland’s late-17th-century maps of Scotland by Timothy Pont (the first maps of Scotland based on personal survey), can be scrutinised by a wider audience than could ever be provided through access to the original.
• added value and increased usability—digital format brings added value which the analogue formats, by the nature of their medium of production, would never allow, such as the ability to enlarge segments, to cut and compare, to print out, or to make comparisons with objects in other institutions or countries.
• faithful copying—digital images do not suffer from use and can be faithfully and precisely copied without damaging the original, and can be distributed relatively inexpensively.
• rapid retrieval—if the associated descriptive database or archive index is adequate, it is possible to locate a particular image quickly from within a large corpus.

7.13 Programmes of digitisation have the potential to play a major conservation role. For example, at the Public Record Office of Northern Ireland (PRONI) the frequent use of the most popular records leads to damage and to the inevitable withdrawal of material for expensive and time-consuming repair. The use of document scanning makes it feasible to offer access to high-quality copies with the added advantage of speedier access to specific parts of an archive. PRONI has hundreds of volumes of valuation revision books—a popular local history and genealogical source—with the information entered in fading inks of different colours. No conservation treatment can halt the fading—only literally closing the volumes can do that—and therefore the only practical option is scanning. PRONI’s experience is typical of that of other institutions: digitisation both increases the possibilities of access and can assist in the preservation of materials.

7.14 By creating vast quantities of digital images of the holdings of museums and galleries it will be possible for visitors to one gallery to examine comparable pieces at another. Comparative views provide a powerful way to help visitors to contextualise what often appear to be single pieces without a context. Over the past twenty-five years the French have built a substantial digital record of the holdings of their museums and art galleries and made hundreds of thousands of these records and associated images accessible over the Internet. With this resource it is possible to see similar works of art which are not displayed together or to use a database of French furniture to identify items.

7.15 Whatever its benefits, digitisation should only in exceptional circumstances be viewed as a preservation medium (i.e. as a substitute for the real object).

Although it does help to preserve fragile and sensitive material by decreasing direct handling, there is still no certainty that digital media can be preserved in the long term (see 10.13-10.17). In the case of moving images and sound the processes of digitisation do not yet capture all the information held in the analogue format. Even the highest commonly available digitisation standard for sound (a sampling rate of 48kHz and 20bits) does not digitise all the information held on analogue tapes in many instances. This means that digitisation is only useful for creating portable and easy-to-use copies. Only where analogue media are at the edge of loss should they be converted to digital form, and even then the question must be asked as to whether or not an analogue copy would not be better. Digitisation projects should be restricted to collections or items of significant national and international importance (e.g. manuscripts, documents) where the benefits of access will justify the expenditure, or those collections which are not currently accessible but contain information important in documenting particular local communities or heritage sector developments.

Recommendations
 
14. The HLF should only support digitisation projects:
 
* where the information content is high and significant proven public and educational benefit would result from the conversion;
* where material is at risk;
* where the benefit of digitisation is greater than the risk to which it subjects the material being digitised;
* where the existing storage medium is no longer suitable as a storage medium (e.g. nitrate film, photographic prints);
* where the methods of digitisation suggest that the digital surrogate might be a preservation medium;
* where digitisation would significantly increase resource accessibility;
* where clear (international) standards will be used to facilitate storage, retrieval, and preservation;
* where ‘rights’ issues (see ¶s 10.11 & 10.12) are not a bar to dissemination;
* as a means of conserving heavily used material;
* when it will provide material of value to the educational sector; and
* when it enhances the ways in which the content of collections can be studied, manipulated, or accessed.
 
15. The HLF should acknowledge that, in the present state of the technology, digitisation should not be seen as a way to create either a preservation or replacement version except under exceptional circumstances (e.g. where the format of the material is not the most useful, such as printed directories).

Information Quality

7.16 There is no point in creating digital resources if their quality, in terms of the accuracy, precision, and comprehensiveness of the data they contain, is wanting. Digital information is culled from a range of sources and the processes involved are not always conducive to the production of high quality information resources. Part of the task of project management is to reduce the risk that the information resources created by projects will have quality deficiencies. For example, manual capture can be prone to subjective biases on the part of the recorder, or errors can creep in during data encoding. The processes of retroconversion of catalogues and finding aids often do not pay sufficient attention to such activities as error checking. There are a number of measures of information quality:

^{[Table Based on:Diane M. Strong, Yang W. Lee, and Richard Y. Wang, 1997, ‘10 Potholes in the Road to Information Quality, Computer, 30(8), 39 and table 1.]}

7.17 Developing quality information resources requires strategic planning. For example, about 70 National Recording Schemes and Study Groups are contributing to National Biological Recording activities of the Biological Recording Centre. They are recording data about fungi, slime moulds, spiders, ladybirds, butterflies, and more than 65 other categories of the biological heritage. Thousands of people are involved in collection. They regularly encounter difficulties in gathering these data and in making certain they are collected to the same standards and data quality. It is the lack of a single coordinating organisational structure, and lack of agreement on standards (¶s 10.18-10.24) which pose major obstacles to the generation of quality data. The need for quality standards in biological recording has been recognised for some time. The Environment Protection Act 1990 which established the Nature Conservancy Councils for England and Scotland and the Countryside Council for Wales, also established a Joint Nature Conservation Committee (JNCC). The JNCC has responsibility, among its other activities, for 'the establishment of common standards throughout Great Britain for the monitoring of nature conservation and for research into nature conservation and the analysis of the resulting information' (Sec 133.2.d).

Recommendations
 
16. In order to ensure that the digital resources created with HLF funding are of a consistently high quality, applicants must detail the active steps they will take to guarantee this quality.
 
17. The HLF should support initiatives to use ICT to bring consistent data collection and inputting standards to the recording of our heritage assets.