Error Analysis in Japanese Writing and Its Implementation in a Computer Assisted Language Learning System on the World Wide Web

Abstract

This article deals with the implementation of error analysis for a Computer Assisted Language Learning (CALL) program using Natural Language Processing (NLP) techniques. This program can be used for learning the passive voice in Japanese on any World Wide Web (WWW) browser. To estimate the range of possible types of errors in Japanese passive constructions, a survey of errors made by foreign students was conducted. The software uses standard NLP tools such as morpheme and syntax analyzers, plus newly developed error analysis and feedback processing components. The program enables learners to type in sentences freely, detects their errors, and displays appropriate feedback messages to guide learners to correct their errors themselves. Evaluation of the program showed that it can obtain a high degree of accuracy and instructional effectiveness.