<![CDATA[The rapid development of artificial intelligence (AI) and multimodal learning has created new opportunities to improve mathematics education, especially in enhancing junior high school students’ engagement and conceptual understanding. Although many studies have explored these approaches, their findings remain scattered. This study conducts a Systematic Literature Review (SLR) to examine trends, implementation characteristics, and the contributions of AI and multimodal learning from 2015 to 2025. Using Publish or Perish (PoP), Scopus, and ERIC, 261 articles were identified, with 30 meeting the inclusion criteria and analyzed using thematic synthesis in Rayyan. The results show that AI integration has progressed from rule-based intelligent tutoring systems to adaptive and generative models that provide rapid feedback, personalize learning paths, and detect misconceptions. Multimodal learning strengthens mathematical understanding through dynamic visualizations, digital manipulatives, gestures, auditory narration, and interactive simulations, supporting cognitive, affective, and behavioral engagement. The combination of AI and multimodal learning yields the strongest effects through hybrid intelligence, merging AI-driven immediate feedback with teacher-led reflection. Overall, both approaches significantly enhance students’ engagement and conceptual understanding, though their effectiveness depends on teacher readiness, technological infrastructure, and pedagogical design. The study underscores the need for improved teacher competencies, better school infrastructure, and curricula supporting adaptive learning, alongside further research on long-term multimodal AI integration in low-resource schools.]]>
Copyrights © 2026
