A semantic proofreading tool for all languages based on a text repository 1
Kai A. Olsen, Bård Indredavik · 2010
A method for finding lexical, syntactic or semantic errors in text in any language is introduced. It is based on a large text repository. The idea is to “follow everybody else”, that is, to compare the sentence offered by the user to the similar sentences in the text repository and suggesting alternative words when appropriate. This concept offers the possibility of taking proofreading further than what is possible with standard spell- and grammar checkers. A prototype for such a system has been developed. This includes a spider that traverses the Web and stores the retrieved text; a builder that creates a convenient index structure for the text repository and the part that analyses the user’s sentence, offering suggestions for improvement.