Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Top-Down vs. Bottom-Up Approaches for Automatic Educational Knowledge Graph Construction in CourseMapper

Qurat Ul Ain, Mohamed Amine Chatti, Amr Shakhshir, Jean Qussa, Rawaa Alatrash, Shoeb Joarder

TL;DR

This work tackles automatic Educational Knowledge Graph construction (EduKG) for MOOCs and compares two pipelines—Top-down and Bottom-up—in the CourseMapper platform. It presents a multi-phase KG pipeline (Text Extraction, Keyphrase Extraction, Concept Identification, Concept Expansion, Concept Weighting) and evaluates the two approaches via a user study and a Simple Random Sampling accuracy assessment. Results indicate that the Bottom-up pipeline achieves higher accuracy and better user experience, though differences are not always statistically significant, and a Human-in-the-Loop is proposed to refine EduKGs before publication. The study offers a scalable framework for EduKG construction in MOOCs that supports personalized learning and adaptive content through learner modeling and learning analytics.

Abstract

The automatic construction of Educational Knowledge Graphs (EduKGs) is crucial for modeling domain knowledge in digital learning environments, particularly in Massive Open Online Courses (MOOCs). However, identifying the most effective approach for constructing accurate EduKGs remains a challenge. This study compares Top-down and Bottom-up approaches for automatic EduKG construction, evaluating their effectiveness in capturing and structuring knowledge concepts from learning materials in our MOOC platform CourseMapper. Through a user study and expert validation using Simple Random Sampling (SRS), results indicate that the Bottom-up approach outperforms the Top-down approach in accurately identifying and mapping key knowledge concepts. To further enhance EduKG accuracy, we integrate a Human-in-the-Loop approach, allowing course moderators to review and refine the EduKG before publication. This structured comparison provides a scalable framework for improving knowledge representation in MOOCs, ultimately supporting more personalized and adaptive learning experiences.

Top-Down vs. Bottom-Up Approaches for Automatic Educational Knowledge Graph Construction in CourseMapper

TL;DR

This work tackles automatic Educational Knowledge Graph construction (EduKG) for MOOCs and compares two pipelines—Top-down and Bottom-up—in the CourseMapper platform. It presents a multi-phase KG pipeline (Text Extraction, Keyphrase Extraction, Concept Identification, Concept Expansion, Concept Weighting) and evaluates the two approaches via a user study and a Simple Random Sampling accuracy assessment. Results indicate that the Bottom-up pipeline achieves higher accuracy and better user experience, though differences are not always statistically significant, and a Human-in-the-Loop is proposed to refine EduKGs before publication. The study offers a scalable framework for EduKG construction in MOOCs that supports personalized learning and adaptive content through learner modeling and learning analytics.

Abstract

The automatic construction of Educational Knowledge Graphs (EduKGs) is crucial for modeling domain knowledge in digital learning environments, particularly in Massive Open Online Courses (MOOCs). However, identifying the most effective approach for constructing accurate EduKGs remains a challenge. This study compares Top-down and Bottom-up approaches for automatic EduKG construction, evaluating their effectiveness in capturing and structuring knowledge concepts from learning materials in our MOOC platform CourseMapper. Through a user study and expert validation using Simple Random Sampling (SRS), results indicate that the Bottom-up approach outperforms the Top-down approach in accurately identifying and mapping key knowledge concepts. To further enhance EduKG accuracy, we integrate a Human-in-the-Loop approach, allowing course moderators to review and refine the EduKG before publication. This structured comparison provides a scalable framework for improving knowledge representation in MOOCs, ultimately supporting more personalized and adaptive learning experiences.
Paper Structure (24 sections, 2 figures, 1 table)

This paper contains 24 sections, 2 figures, 1 table.

Table of Contents

  1. Introduction
  2. CourseMapper
  3. Learning Channels:
  4. Collaboration and Communication:
  5. Awareness:
  6. Educational Knowledge Graphs:
  7. Learner Modeling:
  8. Recommendation:
  9. Learning Analytics:
  10. EduKG Construction Phases
  11. Text Extraction:
  12. Keyphrase Extraction:
  13. Concept Identification:
  14. Concept Expansion:
  15. Concept Weighting:
  16. ...and 9 more sections

Figures (2)

  • Figure 1: An overview of UI of CourseMapper demonstrating different features
  • Figure 2: UI of CourseMapper to Preview and Edit the EduKG