1. Collections and Semantic Thinking
Semantics have always been an integral part of museums. Just as the act of collecting objects into a singular physical site defines museums today and through history, so too does the ongoing act of making meaning with and between these objects. Throughout their (at least) five hundred year story, museums represent the coping strategy of countless historically and culturally differentiated communities to bring some degree of order to the world around them. Traditionally, museums bring fragments of society’s knowledge and experience into a highly controlled environment, a closed system, within which order can then be found – or contrived. It has been the function of museums to find meaning within (or give meaning to) these highly orchestrated aggregated performances. Whether through narrative or systematics, genre or geography, the museum has sought to frame disparate objects with a rationalising discourse into a ‘collection’ or ‘exhibition’. Using numbering, term lists, taxonomy or simply physical proximity, museums have deployed systems that have provided logic, a set of assumptions, and (invariably) a vocabulary to tame and explain their collections. This classificatory project (this syntax) and importantly, the semantics inextricably linked to it, are what has come to define the museum today and through history. Museums, in short, are places where, in a structured way, we give or reinforce meanings to things. This is museum semantics.
More recently (or, at least, in the last forty years) the automated and systematic processing of computer technology has come to support and augment this syntactical and semantic project. Unsurprisingly perhaps, when museums first began to explore the potential of computing it was specifically to the management and documentation of collections that they first turned (Vance, 1973). In contrast to some highly localised, frequently personalised, at times idiosyncratic ordering systems that existed in museums, computing instead promised (indeed, required) a systematized and consistent treatment of collections. Computers could help generate (and, in fact, were dependent upon) a curatorial environment where the semantics of collections could be formalized and standardized. Moreover, the standard classificatory codes, data entry conventions, fixed fields and terminology control that the new automation was reliant upon came to resonate with the professional aspirations of the new managerialism that came to define the sector in the last quarter of the twentieth-century. A professional museum was an ordered museum, and an ordered museum was a museum that used the new documentation – driven by standards and automation (Parry, 2007). Within a generation (or two), therefore, the museum sector became familiar with the concept of a collection being organized and described in a highly structured and systemic way. Standardisation became orthodoxy. At least for classification and management purposes, semantics (the meanings associated with objects) became highly engineered.
However, in the last five years, this ‘engineering’ has been challenged by the new Web 2.0 environment. We are now perhaps all too familiar with the ways in which what was once an extensively broadcast model of centralised, specialised, institutional providers publishing flat on-line pages predicated upon agreed protocols and standards of collections management, has been supplanted by a much more diverse on-line cultural heritage ecology. This is the digital landscape where users become producers, where ‘pages’ are more transitory. This is a landscape characterised by syndication, where the Web (rather than in its old function as repository) becomes ‘event’ and communications space (Pratty, 2007). (Or, to see it in Dickensian terms: this is a move from the Miss Havisham model to the Fagin model. The Miss Havisham model involves the museum sitting there in wedding dress waiting for people to turn up to the party, with cobwebs over the table, having gone to all this trouble, angry and bitter, destined to go up in flames. The Fagin model, on the other hand, is to be more cunning, to have lots of agents running around the city, slipping into all sorts of places, being discovered, not being discovered, moving things around, being exciting and, of course, being prepared to review the situation.) Significantly for our discussion here, this is a domain where meanings become more explicitly contested, where the diversity of on-line activity is matched by the diversity of interpretations and voices (Chun et al., 2006; Chan, 2007); and it is here where the orthodox syntax and semantics of museum collections become both negotiable and negotiated.
2. Collections and Semantic Technologies
Although not always stated as such, technology (specifically computer technology) has had a conspicuous role to play in museum semantics. The early publications of the Museum Data Bank (headed by Robert Chenhall, and based at the University of Arkansas Museum) are testimony to how, even in the early 1970s, museum curators worried about how computer technology would reflect (or shape) the semantics of collections (Vance 1974). Likewise, the first generation of hierarchical database management systems that, in the UK, set the parameters on the MDA data model in the mid-1970s posed a series of questions for the development team on to what extent the terminology and data entered would need to be consistent and controlled (Roberts and Light, 1984). The relational models and systems of the 1980s, and the more flexible standard units of information and documentation protocols that they prompted, saw computing, knowledge capture and documentation influence each other yet again, as demonstrated in the publication of MDA’s SPECTRUM standard in the early 1990s (Nuttall, 2000). Then later in the 1990s, and within the context of the Web, initiatives such as the European Union’s ‘Aquarelle’ project yet again wrestled with how cultural heritage content might be made interoperable despite being contained within different semantic and syntactical systems and within different software and hardware platforms (Dawson 1998). In each case, museums had to reflect not just upon information retrieval, the harvesting of data and the interoperability of their systems, but also (fundamentally) upon how they made meaning with their collections. In each case, these were technologies that aimed to support and reflect museum semantics.
3. The Unresolved Semantic Web
It is therefore into this deeper context of ‘semantic technologies’ that our recent engagements with the Semantic Web might usefully be seen. For five years now the museum sector has had its self-proclaimed ‘primer’ for understanding what might be the core building blocks of the Semantic Web (DigiCULT, 2003). Subsequently, Seamus Ross (2006) articulated a clear and beguiling lexicon and outline of how the Semantic Web can be characterised for the culture heritage sector; namely, that it is participatory (cultural heritage and academic institutions can participate as well as students allowing for growing interconnections of knowledge); interactive (it is possible to trace different pathways, contribute to the process of knowledge development and learn by watching how others are adding to these interconnections); retentive (the Semantic Web has the property of retaining knowledge links and being functional so both people and machines can continue to use them in different ways); trans-disciplinary (in the cultural sector, not only do we think of content as being the curator’s input, but we can also build up interconnections of knowledge stories that other people have and build those into other datasets in other cultural institutions); and trans-cultural (building in the cultural interconnections between, for example, Picasso’s work, when heavily influenced by African art, and the pieces which influenced him, and other cultural activities that were going on at that time). During 2006 and 2007, the ‘UK Museums and the Semantic Web’ project funded by the AHRC also devoted considerable time to exploring the potentials and realities of these technologies for the sector. The objective of this series of workshops (the ‘Semantic Web Thinktank’ as it soon become known) was to identify the key challenges and opportunities that the Semantic Web presented for museums and galleries in the UK and to define how the culture sector should respond to them. The Thinktank brought together experts from a wide range of disciplines, including museum policy and management, cataloguing and computational science. Over nine months, 12 to 15 semantic Web experts and museum professionals gathered in a series of closely documented meetings around the UK to explore ways forward for the sector into this more digitally connected future (http://culturalsemanticweb.wordpress.com/). But more recently still – and more tangibly – the sector has at last begun to glimpse localized, small-scale manifestations of semantic technologies being constructed and applied to real heritage content and contexts (Amin et al 2007).
In each case, these projects continued to recognise the importance of metadata, turning data into meaningful objects sitting within intelligent semantic Web structures. However, as the number of initiatives working in this area grows, a number of recurring obstacles and questions continue to characterise (if not plague) the research and development landscape. At a very fundamental and practical level, as Ross (2006) observed so succinctly, Semantic Web projects continue to be hampered by a lack of standards, a lack of skills, a lack of investment, a lack of strong case studies, and the lack of immediate reward. As well as a series of technical questions (what role for RDF, XML, OWL or Topic Maps?), there remain conflicting visions of how the Semantic Web may play within the cultural heritage sector. Will (should?) the Semantic Web be a unified and collaborative venture within the sector? Or, is it more likely to be (and more likely to succeed as) a series of contained and localized systems and projects linked to a specified set of sites and resources? Likewise, will the ontologies upon which the Semantic Web is so reliant be ‘universal’, inviolate and few in number, or (in contrast) highly localized, variable, liquid and multiple? Is the Semantic Web for the heritage sector likely to grow from a coordinated sector-wide set of agreements, standards and joint initiatives? Or will the sector find itself responding to more user-generated ontologies and community-created solutions for intelligently discovering and searching heritage content?
Within these different versions of a semantic future for museums, there is at one extreme the vision of a hard ‘Semantic Web’, with prescribed and persistent ontologies predicated by the sector on existing collections standards (and perhaps term lists and thesauri) by the professional community of practice itself. On the other hand, there is a vision of a soft ‘semantic Web’, with user-defined (and likely more transitory and disparate) ontologies generated by different communities of interest outside of the institution. Both result in more efficient and intelligent resource discovery, but whereas the former appears to privilege an inward gaze towards more improved collections management, the latter looks instead outwards to how users outside the museum make their own meaning and search for and encounter digital cultural content on-line.
4. The Aspirational Semantic Web
What is the aspiration for the Semantic Web? What would a semantic museum look like, and what problems would it solve? We may be able to find some indication in some of the innovative services and partnerships in which museums are currently engaged.
Recent work by Flickr in partnership with the Library of Congress (http://flickr.com/commons) illustrates the power and potential of working collaboratively with leading technology companies in developing semantic approaches. Using images from the Library of Congress collection as raw material, the popular photo-sharing service is enabling users to 'tag' images with their own descriptive metadata. In addition to providing semantically rich information, the project demonstrates the importance of bringing together content and platform in innovative ways. Politically and economically, this project represents an important step in the process of advocating a Semantic Web to the managers and funders of museum services. The culture sector is able to offer deep, rich contextual data and expertise in mediating and interpreting source materials. The private and commercial sectors are able to leverage technology, finance and market share to bring these cultural resources to a far greater audience than would otherwise be possible.
Experience tends to show that there is a cascade effect among the titans of the on-line world. The interest of Flickr in taxonomy and social classification may also prompt interest from companies such as Microsoft, Amazon or eBay, all of which have strongly vested interests in the successful implementation of semantics as part of service delivery. This cascade may just provide the momentum and technical research and development required to move the Semantic Web on from a conceptual framework to a concrete reality, something which the culture sector on its own is not in a position to achieve.
Most of the Internet's major companies are experimenting with tools, frameworks and systems which enable the user to interact not just with the information, but also at the level of the application itself. Platforms such as Google Base and Yahoo Pipes are encouraging a collaborative approach to experimentation and technical development. In this opening up of the development process, it is becoming increasingly feasible to embed true semantics into the base functionality of systems, and to expose these semantics to a massively distributed network. Above all, then, the semantic museum is likely to be one that is joined-up, integrated into and apparent within existing successful and large-scale consumer services. This, in itself, forms a compelling argument for museums to embrace semantics as a mechanism for reaching a vastly extended marketplace.
In the near future, the museum sector's response to the Semantic Web has to connect to current practice in the creation, management and delivery of digital content. In this context, the traditional standards-based museum digital approach will stand us in good stead. February 2008 discussions on the Museums Computer Group (UK) news list passionately exposed the long-running debate in British museums and galleries about good metadata practice and what it is for. Imagining the real possibility of a socially networked Web built from clusters of networked and semantically connected databases, then the long, long legacy of straight museum data with good URIs, consistent metadata and simple tagging are seen to provide a vitally stable infrastructure on which to build.
Another key development in the current generation of digital services is the erosion of the important of 'place'. As the Web itself moves further towards distributed platforms and dynamic media, it seems likely that Web 'sites' will come to be replaced by Web 'services'. In this world of diffuse identities and services, persistent semantics - the ability to share meanings between platforms as well as records - will become the prime currency. With this likely scenario in mind, we should perhaps be considering the places and contexts where our digital cultural content might be found in future. Already we are familiar with search page results where particles of content or individual collection items are sometimes found. Already we may be familiar with massive numbers of one-page viewers of our sites, visitors attracted by search engines. Better quality search results must surely result in the centre of gravity of the on-line museum running away from the place of publishing towards remote published contexts, like blogs, search results pages, re-skinned or repurposed sites.
In this mid-term Semantic Web world, content needs to be self-describing, imbued with its own meaning and context. On your site it makes sense. When found in Google with a summary, it becomes puzzling. We know we need our content to be trusted, and we want our museum’s name to be considered when the object or text is found in a search.
5. The Disruptive Semantic Web
The genesis of the Semantic Web lies in a very specific technological context. The way it is presented, the jargon it has engendered and the solutions it promotes are strongly geared towards the technologist, and this places the entire momentum of its development at risk. The tendency to this point has been for the technologists to presume that the majority of people need not concern themselves with how the Semantic Web will function, but that they will simply benefit from the better, richer, more meaningful and intelligent experiences they will enjoy as a result. This view, however, contains a fundamental error. It is not possible to ‘build’ a Semantic Web, turn it on and enable people to use it. Instead, the development towards a semantically richer Web depends on the countless organisational and individual decisions made every day across the culture sector. It depends on the needs and expectations of the users of cultural services, and the extent to which semantic technologies and techniques can provide better service to meet them.
Whether one thinks of them as a sector, or an industry, or a public service, museums and related cultural organisations are part of a collective whole which maps in the mind of the ordinary user to the concept of ‘museum’. Preparing museums for a Semantic future is a sector-wide issue which cuts across almost every area – from policy to standards, funding to professional practice. In this sense, the real challenge to pursuing a semantically rich cultural offer lies far away from technology. Effecting large-scale and permanent change across the museum sector depends on a variety of factors. From a top-down view, it depends on Government policy in areas such as technological innovation and social reform. These policy drivers inform both the strategic direction of the sector and, more important, the prioritisation of funding. From a bottom-up view, change depends on champions within organisations with the energy, commitment and management support to reset cultural values and attitudes. It depends on practitioners on the ground recognising and working to address the needs of the people who interact with their collections.
As it is currently presented, the potential of the Semantic Web for museums lacks this quality of immediacy and clarity. Although the high-level, long-term aim is clearly stated, Semantic Web discourse rapidly disintegrates into a multitude of candidate technologies and models, each of which independently lays claim to the orthodox definition of a truly semantic Web. Museums cannot dictate what the Semantic Web becomes: they can only participate in its development. If they are to achieve the change necessary to ensure that they are prepared for the Semantic Web, it will be necessary to lay claim to one particular view of what it is and how it works. If money is to be secured, standards changed and staff employed, then it will be necessary to parcel up the concept of a Semantic Web for museums and quite explicitly sell it – to Government, to agencies, to funders and to the profession.
Furthermore, if, over the last forty years, or even over the last ten years, we didn’t have the time, resources, organisational commitment or systems in place to catalogue our collections, how is that we think we will have any of these things to make the collections semantically rich? The answer, of course, is that we won’t – and this presents a fundamental challenge to the ability of museums even to consider how to participate in a Semantic Web. If the vision for the development of a semantic museums sector is decoupled from the realities and practicalities of making it work, then it quite simply will not happen.
Nor does the development of the culture sector's response to the Semantic Web exist in isolation of the technical development which preceded it, and the current direction this work is taking. Building on the Netful of Jewels report (Keene, S. et al, 1999), much of the past decade of technical development for UK museums has been occupied in pursuit of a vision founded on access and education. This vision depended on achieving a critical mass in the on-line publication of collections information. Broadcast models such as mass-digitisation and the provision of on-line access to museum catalogues were intended to achieve a ubiquity of cultural content that was embedded, coherent and fit for a variety of different purposes. In contrast, the policy focus of recent years has shifted towards making more active use of the capacity that is distributed across the sector. Instead of funding content creation per se, a new generation of funding streams is focusing on developing the infrastructure, standards and skills to aggregate content from multiple sources and to broker it to content platforms and services. The new vision is of a multi-faceted, domain-agnostic system that is capable of using any number of protocols to aggregate data from multiple sources, of repurposing that data, and of expressing it to many different services in structured ways.
As this policy agenda moves forward, however, it becomes increasingly exposed to the risk of diverging from the real, practical and on-the-ground business of managing a collection. The more the emphasis is placed on infrastructure and aggregation, the greater the danger that museums and practitioners will not be able to produce the content, the raw material with which to feed them. In this context, and inasmuch as it represents a next-generation challenge over and above the very real corporate challenges currently faced by cultural organisations, the Semantic Web may very well be a bridge too far. If museums are to become Semantic Web-ready, it will be necessary to divert these current developments to ensure that they address both the technical and cultural challenges surrounding the delivery of semantically rich resources.
In the context of the Semantic Web, and indeed the more general trajectory of technological development in museums, the solution perhaps is prosaic but entirely necessary. Participating in the Semantic Web is no different, in practical terms, from participating in any other type of work. It requires skills and resources, it necessitates organisational and cultural change, and it requires clear practices and protocols. Although the potential of the Semantic Web is clearly transformative, the reality falls far short. In this sense, the progression towards a Semantic Web mirrors that of the development of Artificial Intelligence during the 1970’s, 80’s and 90’s. The end goal is clear, but the path is anything but, and apparently promising approaches often yield very little progress. There is the sense, then, that the Semantic Web – tremendously powerful and promising though it may be – is a distraction from what should be the proper business of a museum in the digital age. Perhaps what is needed is clean and consistent data about objects, structured according to open standards and delivered through open architectures. Perhaps it is more important to develop the skills, both technical and managerial, to ensure that technology is well-planned and implemented, and that it delivers first and foremost against the core objectives of the organisation.
6. The Dilemma of the Semantic Web
At the heart of the discourse about museums and the Semantic Web, then, lies a dilemma.
Intuitively, the imperative of the Semantic Web to connect meaning and object reflects the long-held imperative for museums to define, to classify and to present. There are approaches and models sitting alongside the development of a true Semantic Web which are clearly either parallel to or extensions of existing practices within museums. The Semantic Web also apparently has the potential to resolve some fundamental tensions which are currently inhibiting progress - such as the need to provide better-mediated search and discovery of resources via quasi-intelligent services.
At a conceptual level, therefore, it seems as though the development of semantic technologies and methodologies is the next best step towards a fully integrated, always-on, mediated and contextual on-line cultural offering. The dilemma faced by museums, however, hinges on the significant gap between the vision and the reality of the Semantic Web - a gap which critically undermines the ability of the sector to move forward in a clear and constructive way.
The whole on-line environment is experiencing a second generation that is characterised by rapid growth and equally rapid collapse. From this environment are emerging interesting and challenging models, some of which may point towards a more open and participative future. In the time since the first dotcom collapse in 2001, none of these models has had time to stabilise, to demonstrate its long term effectiveness. None of the Creative Industries has had time to learn about these new models, to test them out or to apply them in truly integrated ways to its services.
Museums are understandably driven to embrace new technologies which support the provision of better services to users. From the earliest days of computing and data manipulation, museums have found ways of harnessing technologies to achieve their aims. Where these technologies have succeeded, it has been because there is a natural and intuitive fit between the functionality of the technology and the requirements of the museum and its users.
The danger of the Semantic Web is that, unlike digitisation or cataloguing, it is not a coherent practice or set of practices. There is no coherent answer to the question, 'How do I do the Semantic Web?' and almost no information with which to make an informed decision about technologies, platforms, models and methodologies. Of course, most well-structured and standardised disciplines didn't start out as such, but the Semantic Web is proving uniquely resistant to specification and definition.
Achieving a constructive role for museums in the Semantic Web demands resources, and securing resources depends on the ability of the sector to express a clear and compelling case for investment. This case ultimately hinges on the extent to which semantics enables more people to reach better museum services. This is the bottom line towards which the Semantic Web for museums must be oriented, and without which it runs the risk of shattering into a thousand disconnected splinters.
As with everything else, the extent to which museums will be able to respond collectively and with confidence to the opportunities and challenges presented by the Semantic Web depends far more on politics, economics and people than it ever did on technology. After all the noise has died down, and the cutting edge of New Technology has moved on, it seems likely that the Semantic Web will prove to be, not our destination, but just another stop along the way.
This project was generously funded as part of the Arts and Humanities Research Council (UK) ‘Research Workshops (Museums and Galleries’ scheme. Special thanks are due to the University of Cambridge, University of Glasgow, University of Leicester, University of Newcastle, 24 Hour Museum (Brighton) and the Victoria and Albert Museum (London) for hosting some of these discussions.
The authors are particularly grateful to all of the participants who formed the Semantic Web Thinktank. They were: Philip Adams, Senior Assistant Librarian (systems Development) De Montfort University; Martin Bazley, E-learning consultant; Stephen Brown, Professor, De Montfort University; Robin Boast, deputy director; Museum of Arch. and Anth., Camb.; David Dawson, Digital Futures, MLA; Dylan Edgar, Scottish Museum Council; Jane Finnis, Director, 24 Hour Museum; Michale Freeman, Curator, Ceredigon Museum, Aberystwyth; Areti Galani, Lecturer, CCHS, University of Newcastle; David Gerrard, Web Developer, De Montfort University; Jeremy Keith, Clear Left Design Ltd.; Brian Kelly, UK Web Focus, UKOLN; Suzanne Keene, Lecturer, UCL; Richard Light, XML Consulant; Frances Lloyd-Baynes, Head of Records, V&A; Mike Lowndes, Interactive Media, Manager, Natural History Museum; Lena Maculan, Digital Heritage Research Group, University of Leicester; Gordon Mckenna, Standards and Systems Manager, MDA; Mayra Ortiz-Williams, Digital Heritage Research Group, University of Leicester; Jeremy Ottevanger, Web Developer, Museum of London; Ross Parry, Digital Heritage Research Group, University of Leicester; Nick Poole, Director, MDA; Jon Pratty, Editor, 24 Hour Museum; Phill Purdy, Development Officer, Museums Library Archives Council; Mia Ridge, Database Developer, Museum of London; Seamus Ross, Director of Humanities, Computing, University of Glasgow; Angelina Roussou Researcher Queensland Institute of Technology; Andy Sawyer, Museum Consultant, MWR Ltd.; Paul Shabajee, Research Fellow, University of Bristol/HP-Labs; Jennifer Trant, Partner and Principal Consultant, Archives and Museum Informatics, Canada; Alex Whitfield, Education Prog. Manager, British Library; Alison Wells, Doc. Assistant, Bristol City Museum and Art Gallery; Dan Zambonini, Technical Director, Box UK.
Amin, A., V. de Boer, L. Hardman et al (2007). “Searching and Annotating Virtual Heritage Collections with Semantic-Web Techniques”. In D. Bearman and J. Trant (eds). Museums and the Web 2007: Proceedings. Toronto: Archives & Museum Informatics, 2004. Consulted February 10, 2008. http://www.archimuse.com/mw2007/abstracts/prg_325000959.html
Chan, S. (2007). “Tagging and Searching: serendipity and museum collection databases”. In J. Trant and D. Bearman (eds). Museums and the Web 2007: Proceedings. Toronto: Archives & Museum Informatics, 2007. Consulted May 25, 2007. http://www.archimuse.com/mw2007/papers/chan/chan.html
Chun, S., R. Cherry, D. Hiwiller, J. Trant and B. Wyman. “Steve.museum: an ongoing experiment in social tagging, folksonomy, and museums”. In J. Trant and D. Bearman (eds). Museums and the Web 2006: Proceedings. Toronto: Archives & Museum Informatics, 2006. Consulted May 25, 2007. http://www.archimuse.com/mw2006/papers/wyman/wyman.html
Dawson, D. (1998). Aquarelle – access to European Culture Heritage. Museums Computer Group Newsletter, spring, 4-5.
DigiCULT (2003). Towards a Semantic Web for Heritage Resources: Thematic Issue 3. DigiCULT project, Information Society Technologies Programme, European Commission. Available: http://www.digicult.info/downloads/ti3_high.pdf
Fielding, R. T.(2000). Architectural Styles and the Design of Network-based Software Architectures. University of California, Irvine. Consulted February 10, 2008. http://www.ics.uci.edu/~fielding/pubs/dissertation/top.htm
Keene, S., B. Royan, D. Anderson (Eds) (1999). A Netful of Jewels: New Museums in the Learning Age. National Museum Directors Conference (NMDC). Consulted February 10, 2008. http://www.nationalmuseums.org.uk/images/publications/netful_of_jewels.pdf
Merriman, N. (2007). A Sustainable Future for Collections. In Delivering Collections for the Future in the North West North West Museums, Libraries and Archives Council (MLA NW). Consulted February 10, 2008. http://www.mlanorthwest.org.uk/museumslibrariesarchives/museums/renaissancenw/ collectionsforthefuture/
Nuttall, A. (2000). History of the MDA. Internal unpublished report. Cambridge: MDA.
Open Archives Initiative Protocol for Metadata Harvesting. Consulted February 10, 2008. http://www.openarchives.org/OAI/openarchivesprotocol.html
Parry, R. (2007). Recoding the Museum: digital heritage and the technologies of change. London and New York: Routledge.
Pratty J. (2006). “The inside out Web museum”. In J. Trant and D. Bearman (eds). Museums and the Web 2006: Proceedings, Toronto: Archives & Museum Informatics. Published March 1, 2006. Consulted February 10, 2008. http://www.archimuse.com/mw2006/papers/pratty/pratty.html
Roberts, D.A. and R. Light (eds) (1984). Microcomputers in Museums. MDA occasional paper 7. Duxford: Museum Documentation Association
Ross, S. (2006). Semantic Web Thinktank: Workshop 1. University of Glasgow, 13 October 2006. Consulted February 10, 2008. http://culturalsemanticweb.wordpress.com/workshop-reports
Simple Object Access Protocol (SOAP). Consulted February 10 2008. http://www.w3schools.com/soap/default.asp
Vance, D.(1973). Computers in the Museum. New York: IBM Corporation.
Vance, D. (1974). What are Data? Museum Data Bank Research Report no. 1. New York: Museum Data Bank Committee.
Z39.50 - International Standard ‘ISO23950’ Information retrieval: Application Service Definition and Protocol Specification’. Consulted February 10, 2008. http://www.loc.gov/z3950/agency/