Data Analytics in e-Learning: Approaches and Applications (Intelligent Systems Reference Library, 220)

✍ Scribed by Marian Cristian Mihăescu (editor)

• Publisher: Springer
• Year: 2022
• Tongue: English
• Leaves: 167
• Category: Library

✦ Synopsis

This book focuses on research and development aspects of building data analytics workflows that address various challenges of e-learning applications.

This book represents a guideline for building a data analysis workflow from scratch. Each chapter presents a step of the entire workflow, starting from an available dataset and continuing with building interpretable models, enhancing models, and tackling aspects of evaluating engagement and usability. The related work shows that many papers have focused on machine learning usage and advancement within e-learning systems. However, limited discussions have been found on presenting a detailed complete roadmap from the raw dataset up to the engagement and usability issues. Practical examples and guidelines are provided for designing and implementing new algorithms that address specific problems or functionalities. This roadmap represents a potential resource for various advances of researchers and practitioners in educational datamining and learning analytics.

✦ Table of Contents

Preface
Contents
Introduction to Data Analytics in e-Learning
1 What is Data Analytics?
1.1 Types of Data Analytics
2 Data Analytics and Learning
3 Limitations of Learning Analytics
4 Future Challenges
5 Conclusions
References
Public Datasets and Data Sources for Educational Data Mining
1 Introduction
2 Related Work: EDM Review Papers
2.1 General EDM Review Papers
2.2 Specific EDM Review Papers
2.3 Findings on EDM Review Papers
3 Review on Other Public Educational Datasets
4 Proposed Methodology for Building Datasets
4.1 The Methodology Used for Data Collection
4.2 Structure of the Dataset
5 Conclusions
References
Building Data Analysis Workflows that Provide Personalized Recommendations for Students
1 Introduction
2 Machine Learning Workflows
3 Inferring Personalized Recommendations by Course Difficulty Prediction and Ranking
4 Personalized Message Recommendation by Usage of Decision Trees
5 Conclusions
References
Building Interpretable Machine Learning Models with Decision Trees
1 Introduction
2 Related Work
2.1 Background Related to View Techniques for Better Model Analysis
2.2 Related Work for Innovative Ways to Rank Instances
2.3 Related Work on Building Interpretable Models
2.4 Weka
3 Design of the Proposed Techniques
3.1 Design of a View Technique for Better Model Analysis
4 Experiments and Results
4.1 Results on the View Technique for Better Model Analysis
4.2 Short Dataset Example
4.3 Validation of the Procedure of Ranking Instances Based on Leaf Analysis
5 Conclusions
References
Enhancing Machine Learning Models by Augmenting New Functionalities
1 Introduction
2 Related Work
2.1 Related Work in Student Modelling Based on Text Analysis
3 Design of Improved User Modelling Based on Messages from E-Learning Platforms
3.1 Algorithm Selection for Data Analysis
3.2 Design of New Functionalities to Improve Student Modelling Based on Forum Activity
4 Conclusions
References
Increasing Engagement in e-Learning Systems
1 Introduction
2 Proposed Approaches for Increasing Engagement
2.1 Modelling Students Based on Their Activity on Social Media Platforms
2.2 Finding the Learners that Simulate Activity and Explore the Correlation Between Social Activity and Learning Performance
2.3 Engagement by Alerts
3 Experiments and Results
3.1 Gathering Data from Several Social Media Platforms
3.2 Marking the Learners that Simulate Activity and Analyze the Correlation Between Social Activity and Learning Performance
3.3 Exploring the Impact of Social Media on Students’ Performance
4 Conclusions
References
Usability Evaluation Roadmap for e-Learning Systems
1 Introduction
2 Related Work
3 Proposed Approaches for Interface Optimisation
3.1 Interface Optimisation by Usability Analysis
3.2 Experiments for Interface Optimisation for Better Usability
3.3 Experiments Obtained from Analysing Key Issues that Influence the Interaction in e-Learning Platforms
3.4 Results Obtained from Exploring How Professors Perceive the Ease of Use of e-Learning Platforms
3.5 Recommending Tutors to Students for Increasing Engagement
4 Conclusions
References
Developing New Algorithms that Suite Specific Application Requirements
1 Introduction
2 Related Work
3 Building a New Classifier
3.1 The General Architecture of the Classifier
3.2 Implementation of the Classification Algorithm
3.3 Visualization Plugin
3.4 Demo of the New Classifier
3.5 Sample Application: Determining Tutors Using the New Classifier
4 Conclusions
References