While the semantic Web remains a twinkle in the eye of Tim Berners-Lee, social media is enabling development of a messier, more human, but perhaps in the end more useful Web ontology. I'm talking about the impact of tagging which is on my mind this morning thanks to two blog posts.
First, David Sifry's post on "The State of Technorati, April 2007″ highlights a move among Technorati users in recent months away from query-based search towards tagged pages as the entry point for surfers interested in subjects.
Let me begin by saying the people we serve and their behaviors on our site have shifted remarkably in recent months. In brief, we've seen phenomenal growth in the use of our tagged media pages. As the use of tags becomes more ubiquitous across all forms of social media and the publishing platforms that support them, they've become the lingua franca of the Live Web - the way in which people all over the world indicate what topics or issues are top of mind and guiding self-expression.
About nine months ago, we began to see a marked increase in the use of those tagged media pages, which back then simply included blog posts and Flickr photos using tags. So, throughout the fall and into December, we introduced a number of improvements and new features to our media pages, including the introduction of a huge range of multiple forms of user-generated content. Today, we include blog posts, photos, videos, podcasts, music, people, and events that share a common tag to give our visitors a view into who's saying what - who's doing what - across the Live Web, all in real time.
...the majority of our page views now are no longer just in real-time keyword or blog search, as would have been the case just six months ago, but also in our tagged media pages.
That's a pretty interesting turn of events. Search, more than any other application, led the growth of Web 1.0. But ironically the more information that becomes searchable, the harder it is for self-directed searches to be specifically useful to any individual. As a result we have witnessed the development the search engine optimization industry, which is helpful to publishers. But also we have witnessed the emergence of social search and tagging which represent a kind of user-directed Web information optimization.
Despite its phenomenal growth, tagging hasn't delivered its full potential in part because there are too many non-interoperable tagging systems and in part because the process of tagging can be so darn inconvenient. That's why I was interested to see Stephen Baker's pointer at BusinessWeek to an academic paper exploring a kind of tag optimization technology. In the paper four researchers at Northwestern describe a system of tag optimization that scans existing tags, ranks them according to effectiveness and suggests the best tags for users to choose. The system, called TagAssist, won the best paper award at the International Conference on Weblogs and Social Media last week.
The semantic Web may still evolve, but it's more likely to evolve as a technological response to the social behavior of surfers than as an abstract technological solution to a theoretical problem. Building actionable functions on top of an ontology that people have already chosen sounds like the right development path to me.
A related piece today comes from Arnaud Fischer, programming director at AOL Search & Directional, who gives an overview of the advantages of social search over traditional search.
link to original post