Social bookmarking services allow users to save and categorise links to webpages to what essentially is a giant public bookmark archive whose elements are accessible from any computer, by anyone. But the technology has a much further applicability than this. This article explains the various applications in detail and explores the implications of social bookmarking for the web as a whole ( download this article as pdf).

The social bookmarking services that we have looked at, such as Delicious or Furl, allow users to save and categorise links to webpages (URLs) to what essentially is a giant public bookmark archive whose elements are accessible from any computer, by anyone. But the technology has a much further applicability than this. In Part II of this series, we look at the implications of social bookmarking for the web as a whole.

If only five people were to use the social bookmarking service you use, it is unlikely that you would come across another user whose interest in, say, Venezuelan soils matches your own enthusiasm for the subject. The converse is also true: The more people that use the service, the more items of interest are you likely to find.

Thus, unlike traditional search engines such as Google, which are static in their utility, the more people who use a social bookmarking service, the richer and stronger the service becomes.

In the early days of the internet, prior to the entrance of Google, many firms, not least of which Yahoo, attempted to develop human-edited hierarchical taxonomies or directories of the internet. Remember those now quite quaint ‘Rough Guide to the Internet’-type books? But the internet was simply expanding too quickly for human editors to keep track of it.

Google transformed this space, making finding items on the internet much easier to find, and human-edited internet directories were largely abandoned. The price was of a loss of precision (the percentage of relevant documents returned) in search. Though Google is comprehensive, it is not precise. A page of great utility to an individual can remain unfound if the document happens not to have many referrals.

‘Folksonomies’

The amateur, or ‘bottom-up’, categorising of items via social bookmarking, on the other hand, over time begins to develop a structure of keywords that define information on the internet, like a self-weaving spider’s web. This informal organisational structure or ‘distributed classification’ is often more simply described as a shared or folk taxonomy, in turn giving rise to the (grotesque) portmanteau, ‘folksonomy’.

This folksonomy potentially could reintroduce the precision to search that was lost in the move away from human-edited directories. The tags or other metadata that users attach to documents is produced by humans, who understand the content of a resource in a way that even the cleverest of programmes cannot. Further, unlike the teams of human editors employed by Yahoo and others to scour the internet to draft directories, attaching metadata does not cost much in the way of effort, time or money.

Thus, given enough time and users, a folksonomy or folksonomies could be used to track and categorise the internet with the precision of earlier human-edited hierarchical taxonomies while at the same time being able to keep up with the sheer size of the internet the way that Google can.

A Google-killer?

According to social bookmarking’s most fervent boosters what was originally conceived of as an online filing cabinet could one day have the unintended effect of challenging Google as the primary navigation device used to traverse the internet.

It should be said, however, that we are a long way off from this yet. According to an academic review of social bookmarking services, as of April 2005, Furl – currently the most popular social bookmarking service – had only c. 100,000 users.

Nonetheless, the giants of the internet have held their fingers to the wind and are all moving into the social bookmarking space. Looksmart, an online search and advertising firm, bought Furl back in September, 2004. Yahoo, which was already the owner of digital photo sharing/tagging service Flickr, bought Delicious in December of last year and will likely use it to replace its own in-house but not terribly successful My Web service that incorporates some social bookmarking elements. In March, Microsoft bought up Onfolio (actually a very powerful web research organisation tool without social bookmarking capabilities, but Microsoft has incorporated it into its new Live toolbar and it is inconceivable that some social bookmarking aspect will not be attached to this product at some point). And the blogosphere was abuzz in February that Google had plans to announce in short order its own social bookmarking service, possibly by tweaking its Search History feature.

In the last year, the growing popularity of social bookmarking has resulted in an explosion of new services, some not-for-profit or academically oriented, others strictly commercial, and each claiming to offer some improvement over other such services.

Social bookmarking ‘clones’

For example, CiteULike and Connotea are academically focussed, with extensive annotation capabilities that go well beyond the more populist and very easy to use Delicious. Furl meanwhile offers a good combination of both tagging and annotation. Mag.nol.ia, which has been described on some blogs as a ‘Delicious-killer’ has a more friendly, more aesthetically pleasing user interface than its competitor, but doesn’t really offer a vastly expanded palette of tools. Technorati’s incorporation of tagging is blog-focussed. Flickr is for images. Wists, sort of a combination of Flickr and Delicious, is online shopping oriented; De.lirio.us is an open source ‘clone’ of Delicious (soon to be consolidated with yet another of these services, Simpy). And then there are Spurl, Kaboodle, Unalog, Listible, Blogmarks, Feedmarker, Backflip, BlinkList, LinkaGoGo, Netvouz, Raw Sugar and on and on (read also this review of various services available).

There may well be a need for niche social bookmarking. The community of academics that use CiteULike, for example, find the sharing of links to various academic journals very useful, while most other users would not have access to journal subscriptions and so the service is of little use to them.

Nevertheless, there is very much a bandwagon effect occurring here, with some developers genuinely interested in improving everyone’s web experience while other, less scrupulous entrepreneurs are just hoping to make a quick buck from advertising attached to a copy-cat bookmarking site.

Anglocentrism

Moreover, readers from NGOs and those involved in development work will immediately recognise just from the names alone the anglocentrism of the above list.

Theoretically, the anglocentrism of social bookmarking should not be a problem for long, for so long as there are sufficient users tagging in a given language, then the service will be useful them as they organically construct their own folksonomy based on their mother tongue separate from but overlapping with the dominant English one.

There is a genuine barrier in that the user interfaces in most of the above are only offered in English. However, there are a number of start-ups around the world providing local-language equivalents, including NTPY (Polish), Fresqui (Spanish), Icio (German), Hatena (Japanese), Kopikol (French), Segnalo (Italian), and Wat Vinden Wij Over (Dutch), and the not terribly user-friendly Brim currently has support for ten languages (including Czech, Hebrew, Norwegian, Colombian Spanish, Brazilian Portuguese and Esperanto. Bulgarian, Russian and Ukrainian are likely soon to follow). A more comprehensive listing of non-English-language social bookmarking services is available on the web.

At the same time, the splintering of different social bookmarking services (whether commercially or, more understandably, linguistically) undermines the very aspect of social bookmarking that

makes the technology so powerful – wherein the more people who tag, the stronger the folksonomy becomes. If taggers are fragmented across dozens of mutually incompatible bookmarking services, the network effect is diminished.

We are at a point in the development of social bookmarking wherein there is a gross crisis of overproduction. There is no need for all these providers. We need consolidation of most or all of the various services (one free service, OnlyWire, is beginning to do just that. OnlyWire allows users to use one bookmarklet for seventeen different social bookmarking services).

Our recommendations

In the meantime, the best services, or rather, for our purposes, those most suited to the needs of those from an NGO or development background would be one of the following three:

• Delicious, for its ease of use and its breadth of users,
• CiteULike, for its extended annotation – although only really of interest to those with access to academic journals, or
• Furl, which combines the simplicity and tagging of Delicious with much of the wider annotative capablility of CiteULike.

The latter is also the most popular social bookmarking service, and is already a favourite with journalists, academics, librarians and teachers.

Before you choose, experiment with a handful of different providers and see which one you feel most comfortable with. Keep in mind, however, that as many of these are small start-ups, the bulk of them inevitably will crash and burn, along with all your treasured bookmarks you saved with them. It’s best to go with a fairly well established service such as Furl or Delicious that have already been acquired by a big league player that is unlikely to go bust.

At the same time, keep you ear to the ground for new developments. Social bookmarking was not initially produced by the plodding behemoths of the internet, but by dedicated individuals who thought they had come up with something ‘cool’, in most cases with little expectation of any financial reward. And it is by these same people that the likely solutions to some of the problems that remain with the technology will be developed.

by Leigh Phillips, science writer Contactivity bv, Leiden

March 2006

download this article as pdf .