Automatic Correction of Definite Article Redundancy Error in the English Compositions of College Students

Lei Wang,Ting Wang

DOI: https://doi.org/10.1007/978-981-13-0008-0_8

2018-01-01

Abstract:Scholars have conducted research on the errors made by language learners in writing both theoretically and practically, and they have made considerable breakthroughs in Second Language Acquisition (SLA) with respect of writing. The learners’ use of articles in English writing displays a certain pattern with factors such as influence from a learner’s mother language, personal idiosyncracy or common misunderstanding. Thus it is feasible to judge whether an article is used correctly in a certain context on a rule basis. This paper adopts a rule-based method, which lets the computer learn how to use articles from empirical sources and automatic correct the Definite Article Redundancy Errors (DAREs) committed by students in their English compositions. With the help of both grammatical and contextual information, the rules in our system are deduced from authentic examples and possess authority and accuracy. The system also allows teachers to add or alter these rules flexibly by either examining the actual cases or updating their professional knowledge as they wish. By doing this, we hope the quality and efficiency of marking compositions in Chinese colleges will be improved.