RE: [xml-dev] Adam Bosworth on XML and W3C

["Bullard, Claude L (Len)" <clbullar@ingr.com>]

It depends on the complexity measure suited to the application domain. 

Approximate entropy is one measure.

"Approximate entropy is a statistical measure to quantify the regularity in
relatively short noisy time series. It is defined as the rate of entropy for
an approximating Markov chain to the process. Useful in deistinguishing
between correlated stochastic processes and composite
deterministic/stochastic models."
http://www.cpm.mmu.ac.uk/~bruce/combib/compref260.html

Rounds complexity is another

"rounds complexity, are defined and then illustrated by designing and
analyzing two algorithms: a parallel summation algorithm which proceeds
along an implicit complete binary tree and a recursive doubling algorithm
which proceeds along a linked list. In both cases replacing global
synchronization with local synchronization yields algorithms with reduced
complexity."
http://csdocs.cs.nyu.edu/Dienst/UI/2.0/Describe/ncstrl.nyu_cs%2FTR1991-539

Kolmogorov complexity is another.  "the length of the shortest effective
description of an individual object" aka, compressability
http://www.cwi.nl/~paulv/kolmogorov.html

Psychological complexity "psychological complexity measure developed at
Clemson University, called the Permitted Interactions (PI) measure, uses
design information to calculate the psychological complexity as a measure of
effort. However there is a general demand for measures that can use
information present at earlier phases. Following this requirement the
measure reported here estimates complexity at the domain analysis phase
which is the earliest development phase in an objectoriented software
process. Psychological complexity relates to the cognitive load imposed on
the developers of the software system which is in turn directly related to
the time to completion of the development process."
http://citeseer.nj.nec.com/167340.html

Structural measures of disorder using graphs (connectance ratio)
http://crl.nmsu.edu/users/madavis/Site/Present/tsld001.htm

and so on.

len


-----Original Message-----
From: Eric van der Vlist [mailto:vdv@dyomedea.com]

How do you measure the complexity of a specification?

Eric (just being curious)