BEGIN:VCALENDAR
PRODID:-//hacksw/handcal//NONSGML v1.0//EN
METHOD:PUBLISH
BEGIN:VEVENT
DESCRIPTION:Click for Latest Location Information: http://semtechbizsf2012.semanticweb.com/sessionPop.cfm?confid=65&proposalid=4552\nSemantic data production often involves multiple ad hoc stages in which\ndata from diverse sources must be converted from relational, XML or\nother formats to RDF, transformed from one vocabulary to another, merged,\nfiltered and inferenced before it is used by its intended applications.\nThis process can become complex, fragile and difficult to maintain as\nthe number and diversity of data sources increases.  Furthermore, when\ndatasets are large it can be difficult to keep them updated efficiently\nas source data changes, especially in a distributed environment.  Small\nchanges can easily result in the wasteful regeneration of entire datasets.\nThis presentation explains how distributed, semantic data production\npipelines can be defined as RDF dependency graphs, automatically driving\ndata production stages based on simple, RESTful HTTP interfaces.\nWrappers enable processing stages to conveniently perform arbitrary\nprocessing while intelligently caching intermediate results and avoiding\nunnecessary dataset regeneration.  This approach makes semantic data\nproduction pipelines easier to describe, visualize, maintain and debug.
DTSTART:20120606T084500
SUMMARY:Automating Semantic Data Production Pipelines
DTEND:20120606T092959
LOCATION: See Description
END:VEVENT
END:VCALENDAR