dajobe: sibling of the Redland RSS 1.0 validator
dajobe: work in progress
dajobe: now reports errors in your N-Triples as well as the triples returned (with lang, XML literals). Should be complete.
dajobe: work in progress
dajobe: now reports errors in your N-Triples as well as the triples returned (with lang, XML literals). Should be complete.
danbri: Implements RDFS + rules using SQL views
danbri: "Tested with MS SQL Server 2000". If anyone here tries it with PostgreSQL etc do share your findings...
danbri: "Tested with MS SQL Server 2000". If anyone here tries it with PostgreSQL etc do share your findings...
DanCon: found via slashdot article
DanCon: gotta study this Bayesian stuff some more...
bijan: Controversy on comp.lang.lisp about whether he was overselling a bit.
bijan: Check out ifile.
bijan: There are some PARC papers on this sort of thing as well, as I recall.
DanCon: google search for bayesian spam] yields some papers from forth.gr etc.
bijan: The key point being that for high accuracy you need a fairly fine grained set of categroies...apparently "spam" and "non-spam" doesn't really cut it (contra the impression Graham gives).