3 Executive Summary

Cite Permalink:
“Having to search sooo many sources – users spend longer finding the sources than they do finding the actual information they are looking for.”
Cite Permalink:
The aim of this short (3-month) scoping study was to determine the feasibility, viability and value of creating an aggregation of metadata about images and time-based media (films and sounds). The research was conducted by EDINA, and is intended to contribute to the implementation of the Resource Discovery Taskforce vision of having a collaborative, aggregated and integrated resource discovery and delivery framework[i] for UK Higher and Further Education.
Cite Permalink:
The scoping study sought to elicit views from a wide range of stakeholder groups: which has revealed a wide range of views, sometimes opposing. Integrating the views of such a range of stakeholders with such diverse experience has been a challenging task.
Cite Permalink:
A total of 80 respondents took part through interviews and online survey: 47 completing the online survey and 40 being interviewed (of which 7 were follow-ups to the online survey). Following the interviews and online survey period, a further 8 people who attended the UK Metadata Forum meeting at the Repository Fringe seminar 2010 participated in a break-out group which further explored the issues addressed during this consultation exercises.
Cite Permalink:
The range of stakeholder groups and range of views was very diverse. In order to extend the consultation exercise of this 3 month study and verify the findings, feedback on this report is encouraged via this blog.
Cite Permalink:
The main conclusions drawn were:
Cite Permalink:
Cite Permalink:
  • There is little that is distinctive about aggregating metadata about images and time-based media, although the variety of implementations of standard metadata profiles leads to complexity in any harmonisation attempt. The conclusions and recommendations arising from the analysis are often not specific to the media format being aggregated, but are related more generally to aggregating metadata.
  • An aggregation of metadata about images and time-based media is useful, only if the purpose and use is clear. This study describes a number of possible uses that indicate the purpose of such an aggregation thus suggesting it is desirable that these metadata be aggregated. It is also inherently valuable to have digital metadata to enable discoverability of related physical resources.
  • To generate potential service provider interest in the aggregation, it may be practical, at least initially, for such an aggregation to focus on aggregating metadata from smaller or lesser-known collections with clear licence terms that are not readily visible to search engines. This would need to be balanced with achieving a sufficient amount of metadata to encourage use.
Cite Permalink:
Cite Permalink:
  • Metadata are particularly important when searching for images and time-based media because search based on full-text-indexing cannot be applied to the resources themselves. Content-based image recognition has not yet developed to the point where textual metadata are not needed.  However, metadata for images and time-based media are often more sparse than for journals and other publications. Among other things, a thumbnail or clip is considered critical to service users but, unfortunately, is often lacking in the metadata. In order to meet user need, collection owners would be required to make thumbnails or video or sound clips available freely and harvestable from their collections.
  • Enrichment of metadata is valuable, especially if this can be automated.  However, if the metadata record is already sparse, it can be more difficult to enrich it by automated means.  Enabling users to enrich metadata with tags (via crowdsourcing) can be beneficial; the new ‘value-added’ metadata should be treated as a commodity that can also be aggregated, which also can be shared.
  • Direct links to the resources described by the metadata should be included in the metadata wherever possible.
Cite Permalink:
Metadata Contributors:
Cite Permalink:
  • The metadata gathered in the aggregation will likely be more useful for applications and hence users if it is ingested in the highest standard and most appropriate format that a collection owner can provide. This will also make management of an aggregation project easier. Ideally collection owners should describe their own content, and self-deposit as much as possible.
  • The level of support required from the aggregator by different metadata contributors may vary according to their level of digital readiness.
  • Sufficient numbers of collections of different types of content, probably around a theme, would be required to gain significant benefit. To support gathering of a useful body of content, the aggregator should evaluate the collections described in collection-level aggregations, such as in IESR, to determine whether the content is suitable for an aggregation of metadata about images and time-based media about individual digital resources.
Cite Permalink:
Metadata Aggregation Model:
Cite Permalink:
  • There is little agreement about whether metadata for images and time-based media should be standardised into a common schema and if so who should do this. The experiences of other aggregators suggest that agreeing a schema would be time consuming. Deploying a common schema based on simple Dublin Core may actually lose information and so may not provide sufficiently useful metadata for images and time-based media.
  • An aggregator should consider the mixed model to centralise the metadata and provide it to others whenever there is a use case to do so.
  • The aggregator should make the aggregated metadata accessible by providing APIs and supporting standard protocols to support developers, and adding to access mechanisms when needed by keeping a watching brief of technology and emerging developer requirements.
Cite Permalink:
Aggregated Metadata Consumers:
Cite Permalink:
  • Aggregations of metadata must provide added value and adhere to open access principles so as to maximise their exposure and encourage service developers and others, such as researchers, to make use of them.
  • It is important to work with service developers to provide the capabilities they need, and thus the metadata to drive the services that their end-users need. The aggregator should also be prepared to provide aggregated metadata to other aggregators such as Europeana in the format of the defined schema.
Cite Permalink:
The report recommends the following:
Cite Permalink:
  • Clearly communicate the benefits of an aggregation with descriptions tailored to each of the different stakeholder types (considering collection owners with different levels of digital readiness separately) in language they understand to encourage participation.
  • Make it as easy as possible for collection owners to contribute metadata: support multiple different ingest formats and protocols due to the varying levels of digital readiness of collection owners. Gather metadata and updates through both harvest and submission, and provide guidance to those who would like to digitise their metadata but have not yet done so. Work with lawyers to develop a process for legal agreement that minimises effort for metadata contributors.
  • Adopt different approaches for collections with different levels of digital readiness, all of which should focus on getting collection owners to describe their own content, and deposit metadata themselves as far as possible, whilst providing guidance when needed. For those who have some technical experience, tools should be provided to enable them to deposit metadata easily, in the best form they can provide.
  • Determine the aggregation model that would be most appropriate for an aggregation of metadata for images and time-based media. Initial indications are that the mixed model may be most appropriate; although further work is required to arrive at a conclusive recommendation.
  • Develop a collections policy in line with end-user needs, and prioritise inclusion of such collections in an aggregation of metadata.
  • Include in the collection policy or contributor agreement a requirement that contributors grant permission to the aggregator to provide publicly accessible or harvestable thumbnail images for all visual resources (images and moving images), and clips for film and sound resources.
  • Engage with service providers early on in development of any aggregation.
Cite Permalink:
[i] Resource Discovery Taskforce Blog
Resource Discovery Taskforce Vision

Total comments on this page:

Comments are closed.