Intelligent Fuzzy Assessment System for English Academic Writing in Engineering Education

Abstract

This paper aims to address the objectivity and reliability of academic writing assessment through the rule based fuzzy logic assessment system and to show the feasibility of computer assisted assessment in language learning. In this study, first, a user-friendly and rule based intelligent fuzzy writing assessment software (IFWAS) system was developed. The designed IFWAS is based on the Fuzzy Group Decision Support System (FGDSS), which provides a score and diagnostic feedback to students. The system has a flexible database to form written reinforcing feedback addressed both to students' weaknesses and strengths in writing. Secondly, to see the efficacy of software in an educational setting, it was implemented to assess the academic English writing skills of 15 engineering students. The focus group interviews conducted with students and teachers after the empirical research underlined the importance of an objective and transparent assessment in learning; thus indicating that the educational use of software had positive pedagogical implications.