E-Learning and Intelligent Planning: Improving Content Personalization

Abstract

Abstract-Combining learning objects is a challenging topic because of its direct application to curriculum generation, tailored to the students' profiles and preferences. Intelligent planning allows us to adapt learning routes (i.e. sequences of learning objects), thus highly improving the personalization of contents, the pedagogical requirements and specific necessities of each student. This paper presents a general and effective approach to extract metadata information from the e-learning contents, a form of reusable learning objects, to generate a planning domain in a simple, automated way. Such a domain is used by an intelligent planner that provides an integrated recommendation system, which adapts, stores and reuses the best learning routes according to the students' profiles and course objectives. If any inconsistency happens during the route execution, e.g. the student fails to pass an assessment test which prevents him/her from continuing the natural course of the route, the system adapts and/or repairs the course to meet the new objectives.

Recomendação de Ações Pedagógicas Utilizando Planejamento Automático e Taxonomia Digital de Bloom

“…Na Tabela 1, tambémé possível observar os elementos utilizados para modelar o perfil estudante. Nos trabalhos apresentados por [Garrido and Onaindia 2013, Garrido and Morales 2014, Limongelli and Sciarrone 2014, Garrido et al 2016, Pireva and Kefalas 2018 o perfil do estudante foi definido com base no estilo de aprendizagem proposto em [Felder et al 1988] em conjunto com outras características, como por exemplo: associado aos desejos e necessidades do estudante; em conjunto com os requisitos do curso, impostos por meio das inter-relações entre OA; ou ainda associadosà TRB. Tais abordagens objetivaram inferir ao estudante uma rota de aprendizagem otimizada em termos de recurso ou de tempo.…”

Course Sequence Recommendation with Course Difficulty Index Using Subset Sum Approximation Algorithms

“…W a n g and Y u a n [21] recommend the courses based on user profiles from users' interest description, browse log and subscriptions. Learning objects are recommended sequentially [22] in which a personalized learning route is suggested to learn the sequence of learning object and if the student fails in the assessment during the learning process, the route will be modified/repaired and recommended with new objectives. The research work proposed by X u, X i n g and V a n d e r S c h a a r [5] recommends personalized course sequence recommendation such that time to graduation is reduced along with an improvement in the student's grades.…”

Effectiveness of Data Mining Approaches to E-Learning System: A Survey

“…Visualization techniques are used where data is in relational form of databases. Due to less interaction between students and instructor, the quality level of E-learning is going downwards [8] presented a survey paper to elaborate Elearning procedure and a way to improve the education. This paper is describing the pros of E-learning and, presenting the challenges during the development of an efficient learning environment.…”

“…This paper is a summarization of previous literature 2010-2016 by use of different data mining techniques, and Inductive Reasoning, Genetic algorithms, Fuzzy logic, Artificial Intelligence methods, Clustering, Visualization methods, Classification methods or Classifiers with Artificial Neural networks [2,6,[8][9][10]. To decrease student's dropout ratio, increase the motivation for students and improve the progress of offered courses with the past experiences of educational systems are the most challenging tasks in [11].…”