Bruce, I've seen citation databases that have flat schemes like this, and it means that you don't know how the individual topics relate to each other. Are they separate views of the document? or does "Russia" give you the location of the "public space"? Frustrating. I would tend to leave them separated into different subject elements (your #1) for purely practical reasons -- that I'm assuming that <subject> will be used in display programs to key the start of a new line, but <topic> might not (I see <topic> as a modifier more than an element -- it could have been defined as an attribute in my mind). But you're right that it's pretty much a coin toss, with no "right" answer. kc On Mon, 2003-11-10 at 10:06, Bruce D'Arcus wrote: > Can someone explain to me how to deal with mapping the following to > MODS? > > <keyword>geography</keyword> > <keyword>public space</keyword> > <keyword>Russia</keyword> > > In others words, we have data that represents "keyword" categories, but > there is only one element; it is totally flat. A tool will do the > transformation, so no human interpretation involved. > > How best to do this? > > Options I see are: > > 1) > > <subject> > <topic>geography</topic> > </subject> > <subject> > <topic>public space</topic> > </subject> > <subject> > <topic>Russia</topic> > </subject> > > 2) > > <subject> > <topic>geography</topic> > <topic>public space</topic> > <topic>Russia</topic> > </subject> > > My hunch is the correct answer is really context-specific, and so we > have to flip a coin. Is that right? > > Bruce -- ------------------------------------- Karen Coyle Digital Library Specialist http://www.kcoyle.net Ph: 510-540-7596 Fax: 510-848-3913 --------------------------------------