Towards an algorithm-based intelligent tutoring system: Computing methods in syntactic management of garden path phenomenon

Abstract

This paper argues that algorithm-based intelligent tutoring systems (ITSs) can give a meaning to an educational content semantically. An answer has been sought to the question of “what should be taught to students when garden path sentence is given to them”. With reference to this aim, a system has been designed to automatically detect the students' cognitive process in decoding the information embedded in garden path sentence. The developed system basically comprises computing schema, linguistic model, flowchart programming and algorithm-based computing methods. The example-based results indicate that garden path sentence has distinguishing features of processing breakdown and backtracking, correspondingly extending the working memory of the subjects. Students' final interpretations of sentences often fail to correspond to the initial input and the theory that comprehension can occur in a “good-enough” manner is supported. As a result, more additional information, semantic and syntactic, should be given in ITSs to the students for the management of garden path phenomenon.