- From: Mike Hogarth <mike@hogarth.org>
- Date: Sun, 04 Mar 2001 12:09:35 -0800
- To: www-xml-query-comments@w3.org
- Message-ID: <200103041209350906.005D0FE9@mail>
Hello: After a long history, terminologies (ie, vocabularies) are becoming increasingly necessary in healthcare for interoperability at the semantic level for decision support, error reduction, aggregated analysis (epidemiology, healthcare system management, etc..). As a result, there is an increasing interest in "terminology servers" as methods for making vocabularies/terminologies available to applications in much the same way that database servers provide data. However, there is no standard "deployment" architecture. The attached paper was presented at the annual medical informatics conference where I proposed a query language as a simple interface for terminology servers. There is significant interest in using a query language approach for vocabularies within the standards body Health Level -7 (HL-7). The issue at hand, however, is that the language assumes an .xml "document" in its syntax. Would you consider changing the syntax slightly such that it allow for "source" rather than "document"? This might allow it to be a more general-use language. I find it to be very close to what is necessary for vocabulary "servers", except the assumption that it is operating on a document --- rather than something more abstract such as a tree+nodes structure. Anyhow, here is an example of what could be done: <xquery-result-set> FOR $a IN distinct(source("diseases)//cardiac) RETURN <concept> <display-name>$a/text()</display-name> </concept> <xquery-result-set> The relevance of terminology/vocabulary servers is not only high in medicine, but becoming more prevalent in other domains. In particular, B2B is now realizing the need for semantic interoperability (ie, exchanging controlled meaning through the use of controlled vocabularies). An example of such an effort are the RosettaNet (www.rosettanet.org) vocabularies. I would urge you to consider adapting your language such that it supporting more abstract "sources" of data are arranged in a tree-node-leaf fashion and not assume a document. Sincerely, Michael Hogarth,MD
Attachments
- application/pdf attachment: TQL22.pdf
Received on Sunday, 4 March 2001 15:11:40 UTC