Introduction

With the rapid expansion of the internet, particularly the world wide web, the increased dependence of users on the information to which it provides access, and the ephemeral nature of websites, greater emphasis is being placed on ensuring the preservation of and longterm access to online content. This review surveys some of the different strategies currently being employed to achieve these goals and introduces some of the principal institutions that have taken up the challenge. For a thorough treatment of the more technical aspects of web archiving, please refer to the preceding chapter by Masanès.

Approaches to web archiving

Many national libraries around the world have been taking an active role in archiving large segments of the web in the interest of preserving cultural heritage. Often, these initiatives are linked with legal deposit requirements that have been updated to include networked digital materials. The first challenge of a web archiving project is defining what to collect. Since the late 1990s, web archiving programmes have tended to adopt one of the following approaches: domain-specific, selective, topic-specific or a combination of the aforementioned. A domain approach involves automatically harvesting websites that make up a particular web space; for example, all sites within the so-called Swedish internet or that fall within the. se domain (more below). The advantage of this approach is that it is comprehensive. The Swedish Royal Library argues that one cannot predict what will be deemed of value to researchers in the future; therefore, resources should be spent on developing the technology to undertake large sweeps of the web and on storage to house what is collected, rather than on the human capital required to make selection decisions (Royal Library, 2005).

The disadvantage of the domain approach is that it does not allow for much control over what is collected. To address this issue, the selective approach prioritizes the online information that is collected; it archives ‘defined portions of Web space or particular kinds of resources according to specified criteria. Selection may be based on the significance or quality of resources, their theme or topic, or by targeting a related set of Web sites’ (PADI, n.d.). The model for this strategy is the National Library of Australia, which started development of its PANDORA archive in 1996 (more below).