Design Objective: Produce Quality Data

csv2rdf4lod-automation is licensed under the [Apache License, Version 2.0](/~https://github.com/timrdf/csv2rdf4lod-automation/wiki/License)

+1 for http://blogs.talis.com/platform-consulting/2011/05/04/triplification/

either make or break the utility of the data represented as triples

Tim Hodson is my hero. Humans matter, and thinking about what we're doing matters.

Our tool is designed to permit quick-and-dirty "like it's source" "intersection" conversions, and it lets us think very hard to make high quality data -- ALL within the same infrastructure. The advantage here is that the process of domain modeling is difficult, but can be vastly improved if we work with concrete examples; namely, the "dumb" conversion. As our understanding of the data increases, we can make model the "without context" data by adding enhancement layers to the "dumber" layers that were created when we didn't know better. This is done by specifying new interpretations of the data using our declaratively RDF vocabulary. In addition, adding that new, awesome representation doesn't hurt anybody that did forge ahead and make an application or do some analysis with the "table intersection" data, because our design monotonically adds descriptions as we learn more about the data. Upgrading is as simple as changing the predicate namespace.

(I'm guessing Tim Dodson the super-contextualness ofhttp://tiree.snipit.org/talis/tables/)

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0018657

http://www.slideshare.net/ldodds/the-rdf-report-card-beyond-the-triple-count