<HTML><BODY style="word-wrap: break-word; -khtml-nbsp-mode: space; -khtml-line-break: after-white-space; "><DIV><DIV><BLOCKQUOTE type="cite"><DIV class="Section1"><P class="MsoNormal"><B><FONT size="2" face="Arial"><SPAN style="font-size:10.0pt; font-family:Arial;font-weight:bold">Cross Application Tag Aggregation</SPAN></FONT></B><FONT face="Arial"><SPAN style="font-family:Arial"> is useful when you want to build on other people's tags. Richard Newman has built a system that does this sort of thing - Richard?</SPAN></FONT></P></DIV></BLOCKQUOTE><DIV>Quite so.</DIV><DIV><BR class="khtml-block-placeholder"></DIV><DIV>I actually wasn't considering specific systems or activities when I tried to formalize my tag ontology[1]. Instead, I tried to precisely model the act of a user's tagging of a resource. It turns out that this kernel is present in every tagging application I've seen, which is convenient! More recently I've applied this ontology to my own software, where it uses tags, and also to integrate my own taggings from del.icio.us, de.lirio.us/Simpy, etc. The details change between services, but there are a great deal of similarities.</DIV><DIV><BR class="khtml-block-placeholder"></DIV><DIV>This integration between services turns out to be one of your use cases: getting data from different tagging services and smushing it together. Once you get them into a shared data model, with shared identifiers — RDF provides these — you can start finding books (Librarything), pictures (Flickr), websites (del.icio.us), and people (my own stuff) which are associated with the tags you want. Pulling in other users' tags introduces some semantic issues, but again lets you do more.</DIV><BR><BLOCKQUOTE type="cite"><DIV class="Section1"><P class="MsoNormal"><FONT face="Arial"><SPAN style="font-family:Arial"><SPAN style=""><FONT class="Apple-style-span" size="4"><SPAN class="Apple-style-span" style="font-size: 13.3333px;"><SPAN style="">Harry Halpin is doing excellent work bridging ontologies to more syntactic formats with the intent to enable use cases such as cross application tagging. We should hear more about this in a separate thread. For here, I would point out that to do this right, you need to join both the tags and the identity information (about the tagger) across applications, communities, or whatever you call it.</SPAN></SPAN></FONT></SPAN></SPAN></FONT></P></DIV></BLOCKQUOTE><DIV>Quite so. Stefano Mazzocchi and I ran into this, and I tend to 'solve' it by using unique URIs for each user's tags. Tough problem; I'd love to hear more about it.</DIV><BR><BLOCKQUOTE type="cite"><DIV class="Section1"></DIV></BLOCKQUOTE></DIV><BR><DIV>Thanks for starting this WG, Tom; it should be interesting.</DIV><DIV><BR class="khtml-block-placeholder"></DIV><DIV>-R</DIV><DIV><BR class="khtml-block-placeholder"></DIV><DIV>[1] <<A href="http://www.holygoat.co.uk/projects/tags/">http://www.holygoat.co.uk/projects/tags/</A>></DIV></DIV><BR></BODY></HTML>