Friday 1 August, 11:00 - 12:30Chair: Vania Dimitrova
Towards Computerized Adaptive Assessment Based on Structured Tasks (page 224)Jozef Tvarozek, Milos Kravcik, and Maria Bielikova
In an attempt to support traditional classroom assessment processes with fully computerized methods, we have developed a method for adaptive assessment suitable for well structured domains with high emphasis on problem solving and capable of robust continuous assessment, potentially encouraging student's achievements, reflective thinking, and creativity. The method selects problems according to the student's demonstrated ability, structured task description schemes allow for a detailed analysis of student's errors, and on-demand generation of task instances facilitates independent student work. We evaluated the proposed method using a software system we had developed in the domain of middle school mathematics.
LS-Plan: An Effective Combination of Dynamic Courseware Generation and Learning Styles in Web-Based Education (page 133)Carla Limongelli, Filippo Sciarrone, and Giulia Vaste
This paper presents LS-Plan, a system capable of providing Educational Hypermedia with adaptation and personalization. The architecture of LS-Plan is based on three main components: the Adaptation Engine, the Planner and the Teacher Assistant. Dynamic course generation is driven by an adaptation algorithm, based on Learning Styles, as defined by Felder-Silverman's model. The Planner, based on Linear Temporal Logic, produces a first Learning Objects Sequence, starting from the student's Cognitive State and Learning Styles, as assessed through pre-navigation tests. During the student's navigation, and on the basis of learning assessments, the adaptation algorithm can propose a new Learning Objects Sequence. In particular, the algorithm can suggest different learning materials either trying to fill possible cognitive gaps or by re-planning a newly adapted Learning Objects Sequence. A first experimental evaluation, performed on a prototype version of the system, has shown encouraging results.
Supporting Users in Creating Pedagogically Sound Personalised Learning Objects (page 52)Aoife Brady, Owen Conlan, Vincent Wade, and Declan Dagger
Successful eLearning is predicated on the application of pedagogies appropriate to online education that respond to the capabilities and needs of the learners. Typically, designing and assembling personalized learning objects that respond to the pedagogical needs of a variety of different learners is an expensive and time-consuming process requiring both domain and educational expertise. Educators have the domain expertise and formal or informal pedagogical knowledge to create quality learning objects. However, they lack the tools and often the specific knowledge of online pedagogical approaches that make it time efficient for them to do so. This paper describes the motivation behind, the workflow supported by and the evaluation of the LO Generator, a tool that offers personalized support and scaffolding for users, who are not necessarily content creation or pedagogical experts, in assembling pedagogically sound personalized learning objects.