ALES: An innovative agent-based learning environment to teach argumentation

摘要

Argumentation is essential in our daily life since we argue all the time in scientific communities, parliaments, courts ... etc. In the field of education, argumentation and argument skills reflect the students' abilities to outline a claim in a logical and convincing way and provide supportable reasons for that claim as well as identifying the often implicit assumptions that underlie the claim. This paper introduces an innovative agent-based ITS teaching environment "ALES", which concerns natural argument analysis. ALES offers two phases; learning phase and evaluation phase. The learning phase encompasses two learning strategies, learning by search and learning by assessment, in which different representative reports that follow the student progress can be produced easily. During learning by search, ALES utilizes mining techniques to expose and retrieve the underlying experts' analyses that are most relevant to the subject of search. Learning by assessment provides guidance through partial and total feedback, which guide the student analysis based on the pre-selected scheme. The evaluation phase aims to assess the student's analysis comparable to the pre-existed expert's analysis. The paper aims to (Ⅰ) describe the constituent models of ALES and their functions, (Ⅱ) present the encompassed tutoring scenarios associated with an illustration of the teaching pedagogy, (Ⅲ) present a comparative study between ALES and other systems in the same field.