Raw

Nice post Danny. Seeing as Uriqr is gradually creeping out into the open, I figured I'd better sort out the performance issues. So, the first step is now completed and Sindice query results are cached, easing the load on their server (one Uriqr query can create many many Sindice queries, especialy for well RDF-d people such as TimBL and danbri) and improving the performance for users of the service.



(In my pre-coffee brain there's an interesting network-based, "rich get richer" effect here, whereby the more well known you are the more often your name is queried, and the greater the chance the results will be in the cache already, making it quicker for people to search for you).



Next step is to update the FOAF data at the backend that powers the initial queries, as we need some fresher data and broader coverage. As Sudarshan points out, if we can extend the underlying dataset to include DBpedia then we automatically get all the URIs in the dbpedia.org URI space, which should help with many famous people. Those of us who are less famous will have to make do with our own URIs, or those generated for us by conferences such as ESWC2006/7 and ISWC2006/7.



So these jobs will get done in the next few weeks I hope. All that's left aside from that is for everyone to make sure they have a URI, and start using them in their FOAF files, especially for the people you know!@en