Abstract:

Learning is a lifelong process, accompanying each individual throughout various requirements of life. This learning path, which is heavily characterised by factors such as learner’s current knowledge, skills and learning styles, is far from linear. This leads to a demand for a learning system that supports learners along their learning paths, providing a personalised and on-demand knowledge delivery experience. Such a system, taking into account individual context, can play an essential role in higher and continuous education, where learning paths tend to shift from heavily guided to an independent and self-moderated learning processes.
In this paper, we present requirements, core concepts and technologies for an enhancement of the Studio educational self-assessment learning system, offering a systematic solution for a personalised self-assessment. The Studio system uses a model of different educational areas, represented as an ontology of knowledge domains. Within this ontology, knowledge areas are stored in relation, capturing the education in the context of their related knowledge. By testing the students on the domain structure, the system detects the learning characteristics of each student as a profile in a complex and developing setting. This approach overcomes static self-assessment tests, which are snapshots, punctuate in time and capturing topics isolated. To cope with developing settings, a test has to have the ability to react and branch in dependency of the performance of a student, adapting “which questions to trigger” to the improving profile of the students. Under this assumption we introduce a strict frame of rules to execute the completion of the profiling and decide how to branch the assessment in dependency.
Additionally, different people may have different learning styles or in other words, different preferred ways to learn, to process and to come to the understanding of new information. This, as the result, may lead to different ways of perceiving, communicating and presenting information. It is also suggested in previous researches, that with similar knowledge tested, learners may perform differently, depending on the format of the test questions and also how the students are asked to reproduce the knowledge. Consequently, in the proposed development, different individual’s learning styles will be taken into account. By adopting an automatic learning styles detection, using Felder-Silverman’s framework, the proposed system will be able to dismiss the environment bias and reach to the individual’s core knowledge. An automatic learning styles detector not only produces methods and procedures for a better adaptive self-assessment system, but also offers a more accurate and timely detection, compared to the current traditional measurement methods of surveying. This detection leads to a personalised self-assessment, using the detected learning profile to customise the self-assessment procedure to the individual student.
Based on the result of the proposed innovative self-assessment system, learning recommendations and a personalised feedback visualisation will be delivered in synergy with the progress of the learner, taking into account the learning characteristics of individuals. This way, the system contributes significantly in supporting the learning process, improving the performance of the student by effectively detecting and closing the knowledge gap of the individual learner.

Keywords:

Adaptive self assessment, automatic learning styles detection, virtual learning environments, educational ontologies, student profiling.