THE RECOGNITION OF IMPERFECT STRINGS GENERATED BY FUZZY CONTEXT-SENSITIVE GRAMMARS

In this paper we extend previous work on the recognition of imperfect strings generated by fuzzy context-free grammars to the recognition of imperfect strings generated by fuzzy context-sensitive grammars. A context-sensitive production rule alphaAgamma --> alphabetagamma has both constitutent (beta) and contexts (alpha, gamma) which can be changed by substituting, deleting or inserting terminal characters. Changed rules derive strings which can be considered to be in the language generated by the original grammar with some certainty. Since some strings may not be recognizable when only one change is made to the grammar, multiple error corrections are described.