On the use of case-based planning for e-learning personalization

Abstract

In this paper we propose myPTutor, a general and effective approach which uses Al planning techniques to create fully tailored learning routes, as sequences of Learning Objects (LOs) that fit the pedagogical and students' requirements.\ud \ud myPTutor has a potential applicability to support e-learning personalization by producing, and automatically solving, a planning model from (and to) e-learning standards in a vast number of real scenarios, from small to medium large e-learning communities. Our experiments demonstrate that we can solve scenarios with large courses and a high number of students. Therefore, it is perfectly valid for schools, high schools and universities, especially if they already use Moodle, on top of which we have implemented myPTutor. It is also of practical significance for repairing unexpected discrepancies (while the students are executing their learning routes) by using a Case-Based Planning adaptation process that reduces the differences between the original and the new route, thus enhancing the learning process

Design and evaluation of planning and mathematical models for generating learning paths

“…Several application areas like intelligent tutoring systems (ITS), 15,16 adaptive e-learning systems, [17][18][19] artificial intelligence (AI) planning and scheduling, 9,11,12,[20][21][22][23][24] learning management systems, 25 ontology-based systems, [26][27][28][29][30] recommender systems 16 , and expert systems 12,31,32 have studied the problem of generating learning paths through learning materials (ie, coursework sequencing).…”

“…Other works consider PDDL-based models and have successfully integrated planning technology into fully operable LMSs. 12,23,25,29 Unfortunately, most of the current planning representations do not consider metric preconditions and effects, as well as metric propagation procedures.…”

“…8 Notice that the design of learning paths could be seen as a planning problem, where learning objectives relate to planning goals and learning activities to actions. [9][10][11][12] The problem is to generate plans that achieve the learning objectives that students must complete from their curricula in compliance with system restrictions and user preferences. We use the Planning Domain Definition Language (PDDL 13 ) to encode our planning models.…”

SMART ITS: um Sistema Tutor Inteligente para Flexibilização e Adaptação de Currículos

“…Trabalhos recentes naárea de STI influenciaram a atual pesquisa pela influência do sistema MATHEMA: [Silva et al 2011] [Garrido et al 2016], que também faz uso de planejamento e, da mesma forma, faz necessário a conversão do plano sequenciado para metadados dos OAs. Contudo, tal planejamento não integra o raciocínio BDI, nem contém suporteà Lógica Temporal, e tampouco possui uma estrutura de curso como apresentada no presente artigo.…”

“…Contudo, tal planejamento não integra o raciocínio BDI, nem contém suporteà Lógica Temporal, e tampouco possui uma estrutura de curso como apresentada no presente artigo. Diferente de [Garrido et al 2016], a proposta apresentada neste artigo visa uma solução genérica que pode integrar suporte a diversas ferramentas para o sequenciamento de OAs por meio do modelo conceitual MATHEMA.…”