Consolidated Moving Image and Sound Database Framework

The UK has a rich heritage of film and radio material. This can provide crucial insights into life during the past 100 and more years. It is a hardly tapped resource for researchers in many disciplines from medicine to fashion, from linguistics to geography.

Increasing quantities of archive television and film content are now becoming available online to the general public (eg Pathe, BBC Archive Collections) or the academic research community (eg. ITN news through the JISC's Newsfilm Online, BUFVC collections, BFI InView). Online catalogues now exist for those collections that can be accessed only by other means, such as archive visits. More will be released as the twin processes of digitisation and copyright negotiations proceed.

However, this content is often prepared and delivered as 'collections'. You have to know where you are likely to find something before you look. There is no one place where you can search for content about a particular subject, locating material from cinema newsfilm, television or radio. This presents a considerable barrier to anyone contemplating using audiovisual content in their research.

This project will build such a place, by producing a common scalable search interface for the nine databases developed and/or curated by BUFVC. These cover material from the earliest of cinema newsreels to next week's television. It would enable a researcher interested in the break-up of Yugoslavia to discover in one place and moment:
- contemporary ITN reports
- the BBC's 'first draft of history' documentary series 'The Death of Yugoslavia' from 1995
- contextual footage of the history of Yugoslavia from newsreel footage of Tito's partisans
- a 10 minute Visnews resume of the background history.
It would allow a geographer to find footage of the same place or the same natural phenomena from the beginnings of cinema to the present; a biographer to discover the many radio and TV interviews given by their subject and any other programmes in which they participated.

At all times, the nature and provenance of the material will be made clear: a newsreel from the 1930s has a very different 'feel' to TV news of the 1960s, as does 1960s news to that of today. Understanding the origin of the content is key to recovering its meanings and then to reusing it for contemporary purposes. The nature of the source collections of content is still vital in this process, as with the equally necessary process of finding content where it is not already available in an online digitised form. Where content is available, an embedded hyperlink will enable it to be previewed as part of the search and discovery process.

The interface will enable researchers to select filters to exclude irrelevant material. It will have a strongly visual design. The design of the interface will be extensively tested with researchers from a wide range of disciplines as part of the project.

The project will create a framework into which further information can and will be inserted when it comes available. Once in place, any relevant information (eg from BBC or overseas archives) can easily be integrated. BUFVC has a 60 year history in providing data on moving images to the academic sector, and is core funded by JISC as well as having over 90% of UK higher education institutions in its membership. BUFVC's continued curation and expansion of this data framework is thus assured.

Two further features will ensure that this database and interface remain a living and growing entity. The software that underpins the work of integrating the existing nine databases in BUFVC's curation will be developed by BUFVC and will be considered for release to the open source community. Second, the Web 2.0 aspects of the interface will allow users to feed in their own information: hyperlinks to specific footage, more detailed metadata and free-text descriptions, contextual material, published academic research work etc.

This project brings together information and resources some of which are available to the general public and the commercial sector, and some which, for copyright reasons, are currently restricted to academic use. Metadata relating to newsreels and TV news, radio broadcasts, Shakespeare etc are universally accessible even when the specific programmes are restricted. BUFVC promotes its service as part of membership to sectors wider than HE/FE. As a charity it demonstrates public benefit by making most activities available freely to the general public. BUFVC have many collaboration agreements with broadcasters, archives and academic partners.

The proposed consolidated database framework will therefore have three kinds of impact.

1. Data available to the public
Unrestricted data has direct impact in its use by researchers seeking archive footage for TV programmes, for museum and other public sector displays, and for the work of charity and voluntary organisations whose moving image usage has been revolutionised in the past decade. Archival moving images are habitually used in public appeals and corporate communications alike, and are a standard feature of many web pages. Every such image has to be found. A single consolidated search interface will be of direct benefit to these sectors, as the current range of discrete catalogues requires considerable foreknowledge of the resources. The single search facility will enable a wider public use driven by the current interest in genealogy, social history and media portrayal. It will significantly enable those undertaking specialist tasks such as reminiscence therapy with the aged (where film is a powerful tool), or architecture firms undertaking restorations (as with the Phoenix Theatre, Charing Cross road, London).

2. Data restricted to academic use
The data restricted to academic use can have wide impact through academic research. A large number of charities commission media-based research on the representation of their particular issue e.g mental health, religion, HIV, cancer. Academic researchers are able to take an overview of material that exists and could be, subject to copyright negotiations, made available for specific uses. We anticipate a growth in the interaction between moving image research in the academy and the growing needs of outside organisations as a result. We also believe that this search facility will drive traffic towards those suppliers (ITN, Pathe etc) who have made their material available for the academic sector.

3. The system itself
The envisaged release of the software under an open-source licence is intended to increase the scope of the database framework. Other users will be able to adopt the system architecture. this will make their data consistent with that within the BUFVC system, and will reduce their costs by elimination of duplicated effort, saving time and avoiding the cost of developing similar systems. The open source model implies that community-driven improvements are fed back to all users of software, potentially creating a loop of self-sustained software improvement.

The BUFVC's aggregation of metadata relating to audiovisual material makes it unusual in the UK context. BUFVC and Prof Ellis are currently being consulted by the BBC over their plans to make large parts of their archive available to the public. The BBC acknowledge that there is much to learn from the academic initiatives that have been brought together within the BUFVC. The BUFVC has submitted a pilot project to the JISC - at the BBC's instigation - to derive a publically useable database from Radio Times listings. This follows the model of the TVTiP listings (lead by Ellis, funded by AHRC)

These impacts will be sustainable. The track record of BUFVC demonstrates its support for projects as ongoing developments. Some services now available have had an onlin