Hi Chris, Muriel, thanks for the suggestion to add schema-specific criteria/indicators and the connection of a dataset to SW-tools, that seems very reasonable and is admittedly missing here. Also, the number of links to the dataset is an interesting indicator.
I've read Felix Naumann's and Richard Y. Wang's thesis and discussed all those other content-related criteria in mine. The reason I discarded criteria like reputation and trustworthiness is because, in my opinion, they only help to improve the perceived quality, not the quality itself. Given two sources A and B, of which A has a good reputation and B has not. If A inadvertently publishes data of low quality, while the data of B is perfectly fine, a consumer might still choose source A, because of its higher reputation. Another criterion relating to the content of data is its accuracy. It is included in my findings, but not as a criterion. Instead, I chose it as a category (content), including criteria influencing the accuracy of data. There is no linkage to language metrics. I only focused on the use of language tags for literal values. I'll have a look at the links sent by Milton. Cheers, Annika -- GMX DSL Doppel-Flat ab 19,99 Euro/mtl.! Jetzt auch mit gratis Notebook-Flat! http://portal.gmx.net/de/go/dsl
