Lecture Structure via Automatic Item Classification

Abstract

In this paper, we present an Automatic Item Classification System, called AICS. This system uses the content structure provided by a teacher to create a content tree, which correlates the items with the contents. The main task of AICS is to classify the items and find the most similar contents associated with the items. Once the relationships between the items and the contents are established, AICS can automatically compute the difficulty of items and examinations. The main contributions of the AICS system described herein are twofold: (1) The system can show the contents that are related to the items and help the teacher quickly understand the difficulty of the examination. (2) The system can provide the contents to the students to help them understand the irrelevant items after an examination.