Development and Improvement of a Corpus-based Web Application to Support Writing Technical Documents in English

Abstract

This paper discusses the development of a Web application that supports non-native speakers of English in writing technical academic documents in English. The application presents authentic English sentences from technical corpora on the basis of calculated similarities to the English input. The updates of this study are 1) highlighting relevant words/phrases of the output sentences and 2) generalizing structures of long output sentences to show the structure more explicitly. Results of experiments are also given to show possible usages and some shortcomings of the application.