[Data-modeling] Modeling datastructures

[Praveen Paritosh]

----- "Jeff Prucher" <jeff at metaweb.com> wrote:

| I think Faye is suggesting that your Big-O values be topics, rather
| than
| text strings, which would enable you to define the variables in the
| description of the type, which should, in turn, reduce the ambiguity
| of the
| expressions.  
| 
| > 
| > That seems like a workable solution, yeah. I still have the 
| > issue of not knowing what variables mean in the expression. 
| > If say something is O(mn), in isolation, that means almost 
| > nothing. When it's O(n), people generally assume n represents 
| > the size of the datastructure in question, so that's not so 
| > ambiguous most of the time, but even then, with things like 
| > Tries, the definition of "size of the datastructure"
| > is rather nebulous.
| > 
| > As an aside, might I suggest a new primitive type for 
| > "mathematical expression"? I could construct one with an epic 

Are you thinking of something like Content MathML [1,2] to model mathematical expressions? You could refer to the properties on that type as arguments and describe functions on them. I agree with Faye's suggestion for its simplicity and ease of use, but I'd be interested if this approach could work to address what if you asking for. 

[1] http://www.w3.org/TR/MathML2/chapter4.html
[2] http://cnx.org/content/m9008/latest/

Thanks, 
Praveen