On the Relative Efficiencies of Context-Free Grammar Recognizers
A number of diverse recognition procedures
that have been proposed for parsing sentences with 
respect to a context-free grammar are described in this
paper by means of a common device.  Each procedure 
is defined by giving an algorithm for obtaining a nondeterministic
Turing Machine recognizer that is 
equivalent to a given context-free grammar.  The formalization
of the Turing Machine has been chosen 
to make possible particularly simple description of
the parsing procedures considered.  An attempt has 
been made to compare recognition efficiencies for the
procedures defined.  For a few simple grammars 
and sentences a formal comparison has been made.  Empirical
comparison of the recognition of more realistic 
programming languages such as LISP and ALGOL has been
made by means of a program which simulates the 
Turing Machine on the Univac M-460 computer.  Several
algorithms for producing grammars equivalent to 
a given context-free grammar have been considered, and
the increase in recognition efficiency they afford 
has been empirically investigated.
CACM May, 1965
Griffiths, T. V.
Petrick, S. R.
