Ever have wondered if there is something around like a command line for the Semantic Web? An online tool that allows you load and query a range of data sources? Well, peradventure Benjamin Nowack has developed a tool, called SPARQLBot, that does all this. It is really great. You launch your favourite IRC client (I use Hydra IRC) and then you (or a SPARQL geek near to you) defines a command that does something useful. Pull in some data, query it, smush it, etc.
Check it out! Play around with it. Experience how to instant-integrate data source on an RDF-basis. You will enjoy it. An example may better illustrate it: Imagine you've joined the #sparqlbot channel at irc:irc.freenode.net; you query for news (from an RSS channel):
mhausenblas>sparqlbot, load <FEEDURI>mhausenblas>sparqlbot, news about XXX
What happens in the background is that the feed gets loaded into the store, and the command triggers a SPARQL query:
SELECT DISTINCT ?item ?o WHERE {
?item a rss:item ; ?p ?o .
FILTER REGEX(?o, "$1", "i")
}
Isn't that great? I experienced it today. It really rocks :) You can add your own commands (given some basic SPARQL knowledge) and extend the bot's intelligence. Not to mention that this nice tool (being based on ARC2) even understands RDFa - very nice, indeed (try also the 'sw faq command'). However, the name SPARQLBot is to modest, IMHO. It is THE Semantic Web command line!
