Automatized Generation of Metadata and Calibrated Test Item Banks for E-learning

Abstract

The development of e-courses entails the description of learning objects (collections of content, practice and assessment items), through the use of metadata, as well as the development of learning content and e-tests, meant to assess the level of knowledge of the learners. An approach for an automated generation of a calibrated bank of test items (that in turn can be used to generate metadata about learning objects) is presented in this paper. The method is based on assessment and accumulation of the answers of learners, teachers, or experts to specific open-type questions in the course of the e-learning. The dynamic calibration of the test items bank is ensured by taking into account concrete parameters for evaluation of test items, and creates an opportunity to design and develop computer adaptive tests. The design and the implementation of the proposed accumulative method entail the establishment and the maintenance of connections between the learning objects and the test questions, through which the generated metadata are extracted.