String Similarity and Misspellings
The problem of programming a computer to determine
whether or not a string of characters is 
a misspelling of a given word was considered.  A numberof
algorithms were evaluated-some proposed by 
other writers, some by the author.  These techniques
were tested on a collection of misspellings made 
by students at various grade levels.  While many of
the methods were clearly unsatisfactory, some gave 
as few as 2.1 percent incorrect determinations.
CACM June, 1967
Alberga, C. N.
