The increasing volume, not to mention variable quality of resources available on the Internet can make searching for that useful resource time consuming and frustrating. Popular search engines such as Yahoo and Alta Vista can deliver hundreds of sources in response to an enquiry, but refined, sophisticated searching is difficult to achieve. Enabling effective and efficient searching is particularly important when you are providing public access to the Internet, where people are often using the Internet for the first time.
Metadata (which is data about data) is being increasingly used in attempts to provide more effective ways of searching and to make Internet resources more accessible. This article introduces an initiative to extend that use of metadata into the public library community.
What is metadata?
Metadata is data that describes the attributes of a resource, e.g. its author(s), title, physical format etc. Two forms of metadata widely used in the library are the simple catalogue card, where the information on the card is metadata about the library item and the more complex MARC record used in OPACs. The importance of metadata in managing information can be grasped by imagining a library without a catalogue - a huge but inaccessible store of information. A useful introduction to the subject is provided by Paul Miller in his article 'Metadata for the masses' .
In the online world, where new resources appear all the time and are often created and maintained by interested individuals rather than large centrally funded organisations, there is an obvious requirement for an extended and enriched use of metadata. Associating a richer form of metadata with resources is a way of enabling search engines to provide more sophisticated searching. Once created this electronic bibliographic description can aid resource discovery, evaluation and effective management of networked information.
There is no one universal format of metadata, different types of resources in different environments require different elements of description. For example, a public library website may only need a brief description, whereas a data archive may require extensive description. One core set of metadata elements that is becoming widely used however, is the Dublin Core .
This is a collection of up to 15 elements of description. Elements in this set include 'resource type' that relates to the category of the resource, such as home page, novel, poem, working paper, technical report, essay, as well as details of the title and creator(s).
Current development work and initiatives
The metadata group at UKOLN is currently participating in a number of projects relating to resource description. One of these is NewsAgent for Libraries , a project in the Electronic Libraries Programme (eLib). The aim is to create an electronic news and current awareness service for library and information staff. Content will include refereed and other papers, reviews and editorial matter from the most highly respected UK journals in the field including Ariadne. By including metadata descriptions of resources used, NewsAgent for Libraries is acting as a metadata demonstrator project.
To support resource description projects tools have been developed to enable metadata to be associated with resources. DC-dot is a web-based tool for automatically generating Dublin Core metadata. Once generated, the metadata can be updated by hand to create HTML <META> tags suitable for embedding into the HTML source of web pages. In conjunction with this, software tools (robots) are being used to automatically gather and index Internet information (harvest).
Public libraries and metadata
If metadata is to have a significant impact on searching and document management these emerging technical solutions need to be coupled with wider recognition and use of metadata beyond the current confines of academic research communities. At present few creators and managers of web resources are providing richer forms of metadata, such as the Dublin Core element set. With so few of these metadata records, few search engines are incorporating the richer metadata elements in to their searching.
To advance the 'metadata cause', into the public library community, UKOLN proposes forming an EARL task group on metadata.
Project EARL (Electronic Access to Resources in Libraries) is a national public libraries Internet initiative . It provides a framework for partnerships within the public and private sectors and actively seeks links with other superhighways initiatives. It aims to make the advantages of the Internet available to all library users and other members of the public. Current task groups within the EARL Consortium are investigating a wide range of subjects including family history, managing access to distributed library resources, training and community information.
This metadata task group will provide a motivating environment within which information providers in the public sector can explore issues relating to the use of metadata with the aim of developing demonstrator services using tools currently being developed.
It is envisaged that the group would run a workshop(s) to disseminate information on current metadata developments, exchange views on issues relating to the use of metadata and decide the best method of sharing these findings with other EARL partners. We would hope to produce guidelines for the use of metadata in the public library environment.
Some of the most pertinent issues that the task group will need to address are:
- Syntax Issues - the meaning and format of metadata labels to be used in the particular context.
- Content Issues - selecting the terms used in keyword lists and deciding on a format for names and email addresses for example. Another related area to consider is granularity. Which resources should be described? In particular, for a multiple-page document does metadata go into every page or just the title page?
- Management Issues - there are a number of ways of creating metadata and managing it across a website. Metadata can be created by the author of an HTML article, by a website manager or by a third party. HTML authoring tools exist to enable authors to embed metadata into HTML tags. Alternatively the metadata about a resource can be held in a database and only embedded into that resource when it is published or served.
Promoting the use of richer metadata in Internet resources can be seen as an example of the public library 'adding value' to information in networked environments; a key future role for public libraries.
If you are interested in contributing to this initiative and your organisation is a member of project EARL I can be contacted at the address given below.
Author detailsSally Criddle,
UKOLN, University of Bath.
Phone: 01225 826250
Fax: 01225 826838