With Michael Kay thinking about tuples in XQuery, I started to wonder about XQuery as a semantic lens?
I say that because in discussions of Linked Data for example, there is always the question of getting data sources to release their data as Linked Data and/or complaints about the nature or quality of the Linked Data released.
While it may not be true in all cases, my operating assumption is that a user wants only some small portion of data from any particular data source. If data is obtained and viewed as linked data or with whatever desired annotations or additional properties, why pester the data owner?
Or to take the other side, why should we be limited by the data owner’s imagination or views about the data? Our “view” of data is probably more valuable to us than its source in most (all?) cases.
I am sure there are cases where conversion or annotation of an entire data set makes analytic, economic or performance sense, assuming you have the resources to make the conversion.
But that won’t be the case for small groups or individuals who want to access large data stores. Being able to query for subsets of data that they can use creatively will be a real advantage for them.
Of course, I am interested in using XQuery to produce input for topic map engines and representing declarations of semantic equivalence.
Suggestions for view as examples?
BTW, when I posted about the XQuery/XPath drafts, I foolishly used the dated URLs. I should have used the latest version URLs. Unless you are tracing comments back to drafts or the history of evolution of the XQuery, the latest version is the one you would want.
XQuery 3.0 – Latest Version Links
XQuery 3.0: An XML Query Language