The Role of Artificial Intelligence in Multimodal Learning Analytics: A Systematic Literature Review

Multimodal Learning Analytics (MMLA) integrates data from various sources (such as sensors, log files, and physiological signals) to comprehensively understand complex learning processes.Artificial Intelligence (AI) and Machine Learning (ML) technologies play an increasingly important role in the field of MMLA due to their ability to process large-scale, high-dimensional data.This systematic literature review aims to systematically organize and analyze the role, applications, main methods, key findings, challenges, and future directions of AI technologies in MMLA research.We conducted a systematic search of relevant literature from the past decade in the Web of Science (WoS) database and found that the core roles of AI in MMLA include: (1) automating the analysis and prediction of learner states and behaviors; (2) providing personalized learning feedback and interventions; (3) integrating and interpreting complex multimodal data streams; and (4) supporting the development and architectural design of MMLA systems.Commonly used AI methods include supervised learning and deep learning, with application scenarios covering collaborative learning, skill development, online learning, and special education.Key challenges involve model generalizability, data noise, ethical privacy issues (FATE), AI interpretability, and the transition from laboratory research to real-world scenarios.