Let's talk about tags, baby. While some continue to debate the usefulness of tag-based folksonomies, others are starting to build abstraction layers on top of a growing body of user-tagged data. Java and flash-based tools are beginning to emerge that visually map the relationships between tags, as well as exploring the relationships between users themselves.

Flickr graph is an application in Flash and Cold Fusion that explores the social relationships inside Flickr. It uses classic attraction-repulsion algorithms to create a realtime graph of users’ contacts, allowing you to visually traverse the Flickr community. This layer of abstraction is made possible by Flickr’s open API, which allows external applications to poll the Flickr user dataset:

In response to a programming challenge by Seb Paquet, Alf at HubLog has developed a visual interface for mapping the relationships between del.icio.us tags:

Users can double-click a tag to foreground that tag and see its tag relationships. Clicking on a small “info” link attached to the foreground tag will provide immediate links to that tag’s page on both del.icio.us and Technorati. This is a powerful visual tool that will look mighty familiar to any users of The Brain mind-mapping software. It helps bring into sharp focus del.icio.us as a tool at the crossroads of social software and knowledge management.

HubLog also provides a visual map interface for examining the relationships between del.icio.us users, based on users’ inbox subscriptions. A ‘subscription’ simply says, ‘please dump into my inbox any links contributed by this user’ or, ‘please dump into my inbox any links contributed by this user that contain this tag.’ Thus, the visualization is a loose representation of ‘interest groups’ — folks who share a common interest in a particular topic as defined by a tag.

Now that Technorati has recently joined on the tag bandwagon (the tagwagon?), there is a lot of buzz in the blogosphere on tagging, taxonomy, folksonomies and distributed classification systems. The debate mostly centers on emergent classification versus top-down hierarchical classification, with proponents heatedly expounding on the virtues of both. What’s your take? Have you begun to use any of the social services that enable tagging? What are the strengths and weaknesses, as you see them?

[links to visual tools via Musings from the Void]