Tools, Techniques, and Applications of Data Visualization in Education and Machine Learning

Data visualization (DV) and learning analytics (LA) play a critical role in simplifying complexity, enhancing interpretation, and supporting evidence-based decision-making across educational, business, healthcare, and policy domains. Despite the rapid growth of research between 2015–2025, gaps remain in methodological transparency, tool adoption, and balanced reporting of advanced visualization techniques. This review systematically examines DV and LA literature to (i) assess trends in publication outputs and geographical contributions, (ii) identify the most frequently applied databases, tools, and visualization techniques, (iii) analyze decision-making outcomes and cognitive load implications, (iv) map target user groups, and (v) highlight persistent challenges and limitations constraining the field. A systematic search of Google Scholar, Scopus, and Web of Science yielded 101,685 initial records. After duplicate removal and screening, 123 studies were included for full analysis. Studies were classified into categories of visualization tools, techniques, application domains, and decision-making outcomes. Descriptive statistics and thematic synthesis were applied, and results are reported with visual summaries. Research outputs show steady growth with peaks in 2021–2024, dominated by journal articles (69.92%) and contributions from the United States (24.39%), China (18.70%), and India (10.57%). The most frequently used databases were Google Scholar (52.03%), Scopus (30.08%), and Web of Science (17.89%). Tool distribution highlighted the dominance of Tableau (44.72%), Power BI (14.63%), and Excel (8.94%), while dashboards (26.83%), bar graphs (16.26%), and line graphs (12.20%) were the most reported visualization techniques. Education (43.09%) and business (39.84%) emerged as the leading domains of application, with decision-making outcomes most often improving business/industry performance (30%) and policy or healthcare (15% each). Cognitive load findings revealed a balance between reduction strategies (25%) and risks of complexity (20%), underscoring design trade-offs. User groups were led by analysts (32.52%), managers (19.51%), and researchers/students (17.07% each). Key limitations included complexity and scalability (20%), interpretability issues (18%), and data integration challenges (15%). The evidence demonstrates that DV and LA provide significant pedagogical, operational, and strategic benefits. However, reliance on dashboards and descriptive methods reflects underutilization of advanced predictive or interactive approaches. Addressing methodological transparency, scalability, and user training will be essential for broader adoption. A framework (Fig. 18) is proposed to integrate inputs, context, methods, mechanisms, users, and boundaries, offering a structured path toward advancing the role of DV and LA in educational decision-making.