Educational data mining and data analysis for optimal learning content management: Applied in moodle for undergraduate engineering studies

Author(s):  
Angelos Charitopoulos ◽  
Maria Rangoussi ◽  
Dimitrios Koulouriotis
Author(s):  
Marta E. Zorrilla ◽  
Diego García

In this chapter we present a BI application delivered as a service on-demand. In particular, it is a data mining service that aims to help instructors involved in distance education to discover their students’ behavior profiles and models about how they navigate and work in their virtual courses offered in Learning Content Management Systems such as Blackboard or Moodle. The main characteristic is that the users do not require data mining knowledge to use the service; they only have to send a data file according to one of the templates provided by the system and request the results. The service carries out the KDD process itself. Furthermore, the service provides an interface based on Web services, which can be called by external software. In short, the chapter talks about the necessity of a service with these characteristics and includes the description of its architecture and its method of operation as well as a discussion about some of the patterns it offers and how these provide instructors valuable knowledge to make decisions.


Data Mining ◽  
2013 ◽  
pp. 1291-1311
Author(s):  
Marta E. Zorrilla ◽  
Diego García

In this chapter we present a BI application delivered as a service on-demand. In particular, it is a data mining service that aims to help instructors involved in distance education to discover their students’ behavior profiles and models about how they navigate and work in their virtual courses offered in Learning Content Management Systems such as Blackboard or Moodle. The main characteristic is that the users do not require data mining knowledge to use the service; they only have to send a data file according to one of the templates provided by the system and request the results. The service carries out the KDD process itself. Furthermore, the service provides an interface based on Web services, which can be called by external software. In short, the chapter talks about the necessity of a service with these characteristics and includes the description of its architecture and its method of operation as well as a discussion about some of the patterns it offers and how these provide instructors valuable knowledge to make decisions.


2019 ◽  
Vol 23 (5) ◽  
pp. 33-43
Author(s):  
Y. Yu. Dyulicheva

The purpose of the paper is the investigation of the modern approaches and prospects for the application of swarm intelligence algorithms for educational data analysis, as well as the possibility of using of ant algorithm modifications for organizing educational content in adaptive systems for conducting project seminars.Materials and methods. The review of the modern articles on the educational data analysis based on swarm intelligence algorithms is provided; the approaches to solving problem of the optimal learning path construction (optimal organization of the learning objects) based on the algorithm and its modifications taking into account the students’ performance in the process of the optimal learning path construction are investigated; the application of particle swarm optimization and its modification based on Roccio algorithm for the reduction of curse dimension in the problem of the auto classifying questions; the application of ant algorithm, bee colony algorithm and bat algorithm for recommender system construction are studied; the prediction of students’ performance based on particle swarm optimization is researched in the article. The modification of ant algorithm for optimal organization of learning objects at projects seminars is proposed.Results. The modern approaches based on swarm intelligence algorithms to problem solving in educational data analysis are investigated. The various approaches to pheromones updating (their evaporation) when building the optimal learning path based on students’ performance data and search of group with “similar" students are studied; the abilities of the hybrid swarm intelligence algorithms for recommendation construction are investigated.Based on the modification of ant algorithm, the approach to the learning content organization at project seminars with individual preferences and students’ level of basic knowledge is proposed. The python classes are developed: the class for statistical data processing; the classfor modifica -tion of ant algorithm, taking into account the current level of knowledge and interest of student in studying a specific topic at the project seminar; the class for optimal sequence of the project seminars ’ topics for students. The developed classes allow creating the adaptive system that helps first year students with a choice of topics of project seminars.Conclusion. According to the results of the study, we can conclude about the effectiveness of swarm intelligence algorithms usage to solve a wide range of tasks connected with learning content and students’ data analysis in the e-learning systems and perspectives to hybrid approaches development based on swarm intelligence algorithms for realizing the adaptive learning systems on the paradigm of “demand learning".The results can be used to automate the organization of learning content during project seminars for the first-year students, when it is important to understand the basic level of knowledge and students’ interest in learning new technologies.


Author(s):  
Sérgio André Ferreira ◽  
António Andrade

A utilização de plataformas tecnológicas com base de funcionamento online, com destaque para os Learning Content Management System(LCMS), tem ganho uma importância crescente nas Instituições de Ensino Superior (IES). Da atividade dos alunos e professores nestas plataformas resulta um imenso trilho de cliques, que se traduz no registo de um enorme volume de dados – Big Data – no sistema. A ideia do Learning Analytics (LA) é simples e tem associado um potencial transformativo muito elevado: o aproveitamento destes dados permite um processo de tomada de decisão mais informada, abrindo as portas a um novo modelo na gestão das IES nos campos pedagógico e da eficiência organizacional. Contudo, a abordagem à temática dos LA ainda está na infância e a operacionalização eficaz exige respostas a grandes desafios no domínio tecnológico, educacional e das políticas. O trabalho aqui apresentado insere-se neste contexto. Na Universidade Católica Portuguesa -Porto está em curso o desenvolvimento de um sistema LA alimentado com dados do LCMS institucional - Blackboard – que tem como objetivo posicionar cada unidade curricular (UC) e faculdade numa matriz de cinco níveis de integração do LCMS no processo formativo. A matriz foi construída com base em modelos internacionais e considerou-se as funcionalidades oferecidas pelo LCMS. Para dar resposta aos requisitos desta matriz, desenhou-se todo o backoffice do sistema de extração e análise de dados no LCMS. Adicionalmente, foi construída e validada uma escala que contempla as mesmas dimensões, para aferição da opinião dos estudantes sobre a integração e a importância do LCMS no seu processo de ensino e aprendizagem. Depois de concluída a construção deste LA é objetivo articular esta informação comos resultados académicos dos estudantes (Sistema de Gestão Académica) e avaliação dos docentes/ disciplinas (SIGIQ) - dando-se passos na construção de um Academic Analytics.


2020 ◽  
Vol 17 (11) ◽  
pp. 5162-5166
Author(s):  
Puninder Kaur ◽  
Amandeep Kaur ◽  
Rajwinder Kaur

In the IT world, predicting the academic performance of the huge student population poses a big challenge. Educational data mining techniques significantly contribute in providing solution to this problem. There are several prediction methods available for data classification and clustering, to extract information and provide accurate results. In this paper, different prediction methodologies are highlighted for the prediction of real-time data analysis of dynamic academic behavior of the students. The main focus is to provide brief knowledge about all data mining techniques and highlight dissimilarities among various methods in order to provide the best results for the students.


2008 ◽  
pp. 107-134 ◽  
Author(s):  
Emma O’Brien ◽  
Timothy Hall ◽  
Kevin Johnson

This chapter looks at the potential to exploit existing technology enhanced learning (TEL) authoring tools to provide customised learning solutions that address both businesses’ needs and employees learning requirements. It examines the feasibility of integrating training needs analysis into existing authoring tools to automate customisation. The chapter outlines a framework for such using best practices in the technology enhanced learning field such as sound instructional design theories, standard compliant metadata sets, and LO granularity while exploiting well established TEL authoring models. The chapter also highlights how this framework was implemented in practice in the form of an electronic tool that ties closely with existing learning content management systems.


Sign in / Sign up

Export Citation Format

Share Document