ConceptThread: Visualizing Threaded Concepts in MOOC Videos

TL;DR

This paper tackles the challenge of rapidly understanding MOOC video content by introducing ConceptThread, a visual analytics system that constructs concept maps and a flow-based thread visualization to reveal concepts and their relationships across MOOC videos. It combines multi-modal data processing (speech transcription, shot recognition, slide analysis) with core NLP and LLM-driven relationship extraction to build a structured knowledge narrative, including four relationship types: Sequence, Association, Similarity, and Inclusion. The approach is implemented as five coordinated views with interactive editing to support the Instructional Hierarchy stages and enables users to explore, compare courses, and refine results, backed by quantitative, case, and user studies showing improved learning outcomes and reduced cognitive load, especially for complex content. The work contributes a novel narrative visualization for MOOC content, automates concept map construction, and demonstrates practical impact for online learning efficiency and knowledge retention.

Abstract

Massive Open Online Courses (MOOCs) platforms are becoming increasingly popular in recent years. Online learners need to watch the whole course video on MOOC platforms to learn the underlying new knowledge, which is often tedious and time-consuming due to the lack of a quick overview of the covered knowledge and their structures. In this paper, we propose ConceptThread, a visual analytics approach to effectively show the concepts and the relations among them to facilitate effective online learning. Specifically, given that the majority of MOOC videos contain slides, we first leverage video processing and speech analysis techniques, including shot recognition, speech recognition and topic modeling, to extract core knowledge concepts and construct the hierarchical and temporal relations among them. Then, by using a metaphor of thread, we present a novel visualization to intuitively display the concepts based on video sequential flow, and enable learners to perform interactive visual exploration of concepts. We conducted a quantitative study, two case studies, and a user study to extensively evaluate ConceptThread. The results demonstrate the effectiveness and usability of ConceptThread in providing online learners with a quick understanding of the knowledge content of MOOC videos.

Figures (10)

• Figure 1: Exploration of Coursera course videos from Econometrics: Random Variables on random variables, with specific information and system configuration displayed in (a). The course's main structure is presented in (c) overview, further divided into layers in (b) to view detailed concepts and their relationships, helping users maximize insight without browsing the video. The corresponding (d) raw data (video) and (e) support panel can further facilitate cognition.
• Figure 2: The system overview of ConceptThread includes a MOOC video processing with four components to extract data and five coordinated views to facilitate interactive exploration of MOOC videos.
• Figure 3: Visual design for concepts and propositions. (a) shows the radial design for Concepts and Propositions with the associated concept, sub-concept, and course style, along with supplementary design utilized in examples and tests; (b) demonstrates the concept presents three key attributes: duration, importance, and time distribution.
• Figure 4: The visual encoding of the four basic relationships.
• Figure 5: Two alternative designs of Concepts and Propositions