Below are a couple of (in some cases perhaps a bit wild) ideas that I would like to implement (or at least give a try ) in coming versions of SNOB.

• improved subset editor
  That shouldn’t too hard, it will just cost time and some input from end users. So start hammering that keyboard, because this is probably the least exciting idea.
• integrate a description logic classifier
  That wouldn’t be too hard either, given previous experience in developing/implementing a GALEN Terminology Server. This would allow you to find misclassified concepts, but also where to put new concepts.
• separate the back- and frontend into a client/server architecture
  This is probably the easiest to do, since the source of SNOB is already very modular. The client and server part are already separated in two separate libraries, all I have to do is put an interface on the server part. The risk is of course that people out there start building their own clients on top of the server and don’t ask me to build it for them.
• API to integrate SNOB into your own applications
  What form should such an API have? ActiveX/COM, TCP/IP socket,...

If you’ve got any other (wild) ideas, just put them on this page. Or if you want to press me in doing one of the above, let me know.

• natural language generator to support (semi-)automatic translation
  Again, the GALEN projects has given me lots of experience in using a natural language generator. But even without such a thing, it is my strong believe that a first provisional translation of the complete contents of SNOMED CT can be done in about 1-2 weeks. At least for me this is the most exciting idea, and thus on the top of my list. I might try to have a first provisional translation of the complete SNOMED CT contents in Dutch before the end of March.