Introduction

It is clear that people will use metadata to find, identify, select and obtain information resources, and to navigate collections of information resources, for a long time to come. When we consider the future of metadata, then, the question is not whether there is a future, but what that future might look like. There are several aspects to this question, associated with aspects of the metadata we have discussed in the preceding chapters. To what extent will metadata be used in future information retrieval systems, given the rise of content-based retrieval? What kind of metadata will be needed in future information environments? Who will be creating and managing this metadata? What value will we place on it? How will metadata be shared across systems? Will metadata become more, or less, standardized? All these questions are interrelated, of course, and can be answered only by considering the future of the information environment as a whole.

Three approaches

In today's online world, access to information resources can be provided by taking one or more of three general approaches. First, there is the approach which bypasses metadata altogether and uses computers to analyse the content of information resources to retrieve those most likely to be relevant to users’ needs. Second, there is the ‘Web 2.0’ approach, in which end-users, contributors and authors provide the metadata. Third, there is the ‘traditional’ approach, in which resources are organized and described by information professionals, in the ways that they deem most effective for their clients. We can address the questions posed above by examining the prospects for each of these approaches.

Ultimately, approaches to information retrieval reflect different perspectives on the nature of information resources themselves. In the first chapter, we observed that every information resource contains a message. This can be viewed ‘objectively’, that is, as an independent object, and analysed by a third party without reference to either creator or recipient. Although the third party could be a person, such as an information professional, this perspective is most clearly adopted by content-based retrieval systems. Alternatively, the message can be viewed and analysed according to the meaning intended (or thought to be intended) by its creator. This perspective is reflected in metadata supplied by authors and publishers. Again, the message can be viewed and analysed according to its utility, a perspective frequently associated with user-generated metadata.