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. CA670510 JB February 28, 1978 10:28 AM 1112 4 1592 1592 4 1592 1592 4 1592 1592 4 1592 2111 4 1592 2111 4 1592 2339 4 1592 2339 4 1592 440 4 1592 982 5 1592 1112 5 1592 1592 5 1592 1592 5 1592 1592 5 1592 2111 5 1592 2339 5 1592 619 5 1592 830 6 1592 1112 6 1592 1112 6 1592 1592 6 1592 1592 6 1592 440 6 1592 619 6 1592 619 6 1592