Data Fusion in Classroom-Based Multimodal Learning Analytics: A Systematic Literature Review

Yuxuan Wang, Xiaoqing Gu

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Scopus citations

Abstract

In classroom environments, this paper presents a systematic review of Multimodal Learning Analytics (MMLA) following PRISMA guidelines. It explores the relationship between learning indicators, multimodal data types, and data fusion strategies to improve MMLA's decision-making efficacy. The main findings include: 1) Complex mappings exist between learning indicators and multimodal data, including within learning indicator layers; 2) Engagement stands out as the foremost learning indicator; 3) Video data predominates in MMLA applications; 4) Feature-level fusion is the leading strategy for data integration. The study also highlights both theoretical and technological challenges, emphasizing the vital influence of learning theories.

Original languageEnglish
Title of host publicationISLS Annual Meeting 2024
Subtitle of host publicationLearning as a Cornerstone of Healing, Resilience, and Community - 18th International Conference of the Learning Sciences, ICLS 2024 - Proceedings
EditorsRobb Lindgren, Tutaleni Asino, Eleni A. Kyza, Chee-Kit Looi, D. Teo Keifert, Enrique Suarez
PublisherInternational Society of the Learning Sciences (ISLS)
Pages951-954
Number of pages4
ISBN (Electronic)9798990698000
DOIs
StatePublished - 2024
Event18th International Conference of the Learning Sciences, ICLS 2024 - Buffalo, United States
Duration: 10 Jun 202414 Jun 2024

Publication series

NameProceedings of International Conference of the Learning Sciences, ICLS
ISSN (Print)1814-9316

Conference

Conference18th International Conference of the Learning Sciences, ICLS 2024
Country/TerritoryUnited States
CityBuffalo
Period10/06/2414/06/24

Fingerprint

Dive into the research topics of 'Data Fusion in Classroom-Based Multimodal Learning Analytics: A Systematic Literature Review'. Together they form a unique fingerprint.

Cite this