Personalized Spell Checking using Neural Networks

Spell checkers are one of the most widely recognized and heavily employed features of word processing applications in existence today. This remains true despite the many problems inherent in the spell checking methods employed by all modern spell checkers. In this paper we present a proof-ofconcept spell checking system that is able to intrinsically avoid many of these problems. In particular, it is the actual corrections performed by the typist that provides the basis for error detection. These corrections are used to train a feed-forward neural network so that if the same error is remade, the network can flag the offending word as a possible error. Since these corrections are the observations of a single typist’s behavior, a spell checker employing this system is essentially specific to the typist that made the corrections. A discussion of the benefits and deficits of the system is presented with the conclusion that the system is most effective as a supplement to current spell checking methods.