Information Theory Applied to the Conversion
of Decision Tables to Computer Programs
Using ideas from information theory, this paper
develops a heuristic algorithm that converts 
a limited entry decision table to a tree structured computer
program with near minimum average processing 
time.  The method is applicable to any limited entry
decision table and does not require that actions 
have single rules or that the cost of testing conditions
be equal.  It is thus more general than the 
previously published heuristic algorithms.  Compared
to the optimal algorithm of Reinwald and Soland, 
this algorithm is easy to code and takes a much smaller
translation time; it is thus felt that it is 
more useful in practice.  The algorithm is well suited for
manual conversion of decision tables to flowcharts.
CACM September, 1973
Ganapathy, S.
Rajaraman, V.
