Scribd
Upload
19 views17 pages

Adaptive Digital Learning - Personalised Learning

Uploaded by

rkhumara21
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
19 views17 pages

Adaptive Digital Learning - Personalised Learning

Uploaded by

rkhumara21
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 17

DESIGN AND IMPLEMENTATION OF ADAPTIVE

DIGITAL LEARNING FRAMEWORK FOR MATHEMATICS


IN PRIMARY SCHOOLS
1. Research Background

• Mathematics is an essential subject in primary education in Malaysia, and it is crucial


to ensure that students develop the necessary skills and knowledge in this subject
• The need for adaptive digital learning arises from the fact that traditional classroom
teaching often fails to cater to the individual learning needs of students especially in
subjects such as science and mathematics as captured by Programme for International
Student Assessment (PISA) 2018.
• E-learning through online platforms such as Google Meet and Zoom has become a
universal learning way. E-learning allows the study to be done in a non-face-to-face
manner, but it makes teachers hard to capture the standards of every student.
• Technology-based adaptive learning systems is a personalised learning system which
can provide students with learning assistance, study materials based on specific needs
and feedback according to their performance.

• The development of an adaptive learning framework for mathematics subject for


primary school students in Malaysia has the potential to improve student engagement
and performance in mathematics, which is essential for their future academic and
professional success.
2. Problem Statement

• Traditional approach to teaching mathematics may not be effective in meeting the


diverse learning needs of primary school students.

• Students come from different backgrounds, have different learning styles, and vary
ing levels of prior knowledge and skills.

• Personalised learning is not provided in currently existing learning management sy


stem such as Frog VLE and Google Classroom.
•
• High possibilities that weak students will be ignored since it is hard for a teacher
to monitor every student through online platform.

• Leads to increased school drop out rate for 2020 according to MOE. 356,438
for primary schools in Malaysia.
2. Problem Statement

Malaysian Student’s Programme for International Student Assessment (PISA 2018)


statistics.

• On average, 15-year-olds score


440 points in mathematics
compared to an average of 489
points in OECD countries.

• Index of students’ cognitive


adaptability in Malaysia was one
of the lowest among PISA-
participating countries and
economies. (-0.3 PISA Index,
rank 60/64 , 2018).
3. Objectives

1) To identify the current challenges and limitations in mathematics learning at


primary school level in Malaysia.
2) To examine the effectiveness of adaptive learning techniques and their potential
in enhancing the learning outcomes of primary school students in mathematics.
3) To design and develop an adaptive learning framework for mathematics subject
for Malaysia primary school by integrating suitable technique and method that
can cater to the diverse learning needs of students.
4) To evaluate the effectiveness of the proposed adaptive learning framework
through pilot testing and comparison with traditional teaching methods.
4. Expected Research Contributions

i) A comprehensive understanding of the challenges and limitations of mathematics


learning at primary school level in Malaysia.
ii) An evaluation of the effectiveness of adaptive learning techniques in enhancing
the learning outcomes of primary school students in mathematics.
iii) The development of an adaptive learning framework for mathematics subject for
Malaysia primary school that aligned with the Malaysia curriculum.
iv) An enhancement for current adaptive learning algorithm.
5. Literature Review

Learning Style Models

• Learning style models used in adaptive systems include Felder-Silverman, VARK, Honey
and Mumford, Gardner’s Theory of Multiple Intelligences, and Entwistle’s Theory.
• Additionally, multiple techniques have been used to predict the learner style of users,
such as Questionnaires, Fuzzy Logic, Evolutionary Algorithms, Artificial Neural
Networks, Decision Trees, Statistics, Content-based Filtering
5. Literature Review

Learner Model Factors in Adaptive Systems


• The learner model is composed of a multitude of user traits. Such factors are the
following: learning style, prior knowledge, learning goals, cognitive style, personality
traits, learner’s device, learner tracking, and working memory capacity.
• Furthermore, it is of great importance the type of model. Namely, if the model is
implicit, it is constantly updating based on the usage data or explicit, meaning that the
model is static.
5. Literature Review

Learner Model Factors in Adaptive Systems


5. Literature Review

Adaptation Elements and Techniques in E-learning Systems


Elements have been classified into content, navigation, presentation, and assessment.
Adaptation Techniques are Rule-Based, Evolutionary Algorithms, Intelligent Agents,
Decision Tree, k-NN, Machine Learning, Recommendation System, and Fuzzy Logic.
5. Literature Review

User Groups
• User groups that participated in empirical studies are learners, learners and teachers,
high education experts. Educational levels are Students of Higher education, Secondary
education, and Primary education.

• The scientific domains where the adaptive e-learning systems were tested are Computer
Science & Engineering, Mathematics, Social sciences, Chemistry, English, Earth Science
, and various disciplines.

• Finally, concerning the Educational and ICT Impacts that referred in the literature are:
Higher learning performance or Positive impact on learning outcomes facilitates learning
processes, enhance learning satisfaction, effective communication with teachers,
enhancing the learning self-efficacy, suitability of learning materials, improved
algorithms/techniques, improved data mining results.
5. Literature Review

User Groups
6. Research Methodology

Display of Granularity of Learning Objects

Course is being contained from several modules and each module contains certain number of
lessons. Lessons are collection of learning objects which are suitable for the topic of the less
on. Each learning object is contained from the text (several paragraphs), files which include
additional content and multimedia content (audio files, video files, images, clips).
6. Research Methodology

Adaptive Learning Process

Student

The learning process consists of learning objects


and additional activities.

Learning Objects (L01) are entities that deal with


Student
a topic and consist of sections.

As the main activity for checking learner’s knowl


Student
edge within the adaptive learning process.

Student
6. Research Methodology

Adaption Engine (Algorithms)

Adaptation engine (algorithms) use students’ resu


lts which updates after each work on a task and in
accordance with those data generates the path of
learning process.

This iteration of validation, at the end of each LO


, serves to check the newly acquired knowledge
of the student and their progress after reading the
teaching materials.

The student’s progress is measured by comparing


the results obtained at the initial and final task
within a single learning object.
6. Research Methodology

Algorithms for selecting sections that will be displayed in LO


• Algorithm for selecting sections within a learning object is based on a test group and student models.
• Before a new student starts a lesson, the system proposes an input test. Based on the answers and the
knowledge demonstrated by the student at the initial test, the system creates and stores data within their
student model.
• A testing_pre_L01 vector is created which relates to the test results before a learning object 1.
• Now it is necessary to determine which sections will be shown to the current student. Unlike the tester, this
student will not get all the sections and related questions. The algorithm proposed will uses the k-nearest
neighbors’ algorithm for selection of questions and level of difficulty.
• The closest K students are then considered for creating a learning path for student who has just done the test.
• After the algorithm decides which sections to be shown to the new student, a binary vector is generated,
decision tree view sections.
• If the value of the element is set to 1, this means that the section with the same ordinal number will be
displayed, and if the value of the element is set to 0, this means that the section will not be displayed.
6. Research Methodology

System Data Flow Diagram

Generating
Questions
Questions
Pool

Solving
Questions

You might also like