Applying Turing’s Ideas to Search – Boxes and Arrows: The design behind the design applies the Turing test to the problem of understanding searches in order to provide better results. Ferrara suggests we need to revisit the parsing approach (moving on from the pattern-matching paradigm) and to develop “social ontologies” in order to get better search results. The “social ontologies” are – if I have understood correctly – wikis of relationships that can then be accessed by search engines to make semantic inferences. The ontologies would have to be socially constructed as there is just too much information out there to put it all together any other way. It struck me that this is a bit like what SKOS is essentially hoping to do. Once upon a time I wanted to build a fully linked thesaurus of the English language where every word was linked to every related word, so you could navigate through the entire language, following pathways of meaning, with no word left out. People thought it was a daft idea, but compared with trying to build ontologies of everything, it doesn’t seem so crazy. Just shows how times have changed!