The real crux of the issue is designing a (hopefully dynamic) schema that can be used to "contain" and catalog information.  My specific goal is to create a format where additional knowledge (from other sources like wordnet, Cyc, whereever... ) can also be aggregated in.  But the initial challenge is how to capture the very rich and complex "meanings" of words by building a "system" within which they live.   This challenge is the continuing pursuit of this goal